Tenable Blog

The cost of ransomware downtime in schools gets pegged at $500K-plus per day. Meanwhile, check out the AI-usage risks threatening banks’ cyber resilience. Plus, Uncle Sam is warning about a dangerous Iran-backed hacking group. And get the latest on AI-system inventories, the APT29 nation-state attacker and digital identity security!

Dive into six things that are top of mind for the week ending August 30.

1 - Ransomware downtime costs to schools: About half a million dollars per day

After suffering a ransomware attack, schools and colleges lose an average of $550,000 per day of downtime, a heavy financial burden considering they remain offline for an average of about 11 days.

That’s according to an analysis by research firm Comparitech of almost 500 ransomware attacks against U.S. educational institutions between 2018 and this year.

“Most schools faced astronomical recovery costs as they tried to restore computers, recover data, and shore up their systems to prevent future attacks,” reads a Comparitech blog about the research published this week.

(Source: Comparitech, August 2024)

Highlights from the research include:

• In 2023, the average downtime suffered by an educational institution due to ransomware was 12.6 days, up from almost 9 days in 2021.
• The 491 ransomware attacks analyzed between 2018 and 2024 affected about 8,000 schools and colleges
• The average ransom demanded was $1.4 million.
• Collectively, these 491 incidents cost about $2.5 billion and involved the breach of 6.7 million individual records.

For more information about cybersecurity threats to educational institutions:

• “Cybersecurity Preparedness for K-12 Schools and Institutions of Higher Education” (U.S. Department of Education)
• “$200 Million Cybersecurity Funding Available for K-12 Schools and Libraries through FCC Cybersecurity Pilot Program” (Tenable)
• “Survey Reveals Top Cybersecurity Issues in Education” (EdTech Magazine)
• “School cybersecurity threats on the rise” (K-12 Dive)
• “Yes, University Cybersecurity Is Still a Concern” (Inside Higher Ed)

2 - GenAI risks among banks’ top cloud security challenges

As financial institutions aim to boost their cloud security, they face a variety of obstacles, including data-privacy and data-integrity risks posed by their use of generative AI systems.

That’s one takeaway from the Cloud Security Alliance’s upcoming report “Cloud Resiliency in Financial Services,” based on a survey of about 860 security pros, CISOs and financial services leaders.

Other AI-related concerns cited by respondents include the danger of suffering AI-boosted cyberattacks, as well as issues with data accuracy, information bias and regulatory compliance, according to a CSA blog posted this week.

 

 

Meanwhile, respondents’ top-three challenges with cloud service providers are:

• Cloud settings misconfigurations, cited by 62%
• Integration of cloud and third-party services (52%)
• Effective identity and access management systems (35%)

Other barriers cited by respondents include a lack of qualified cloud-security staff and difficulties with serverless and containerized environments.

More key findings include:

• The NIST Cybersecurity Framework has been adopted by 67% of respondents. The ISO/IEC 27001 standard for information security management is also popular.
• Seeking simplicity and easier management, a majority of respondents (78%) prefer to use a single cloud services provider.
• Enhancing disaster-recovery preparedness (60%) and infrastructure scalability and availability (58%) are high-ranking priorities.

To get more details, read the CSA blog “The State of Cyber Resiliency in Financial Services.”

For more information about cybersecurity trends in the financial sector:

• “Is AI Making Banking Safer or Just More Complicated?” (Bank Infosecurity)
• “Seven cybersecurity threats for banks in 2024—and some smart precautions” (ABA Banking Journal)
• “Cybersecurity remains number one risk for global banks, as financial risk moves back up the agenda” (Institute of International Finance)
• “Ransomware in the financial sector: What to know and how to respond” (ABA Banking Journal)
• “Global financial stability at risk due to cyber threats” (World Economic Forum)

3 - CISA: Iranian hackers unleash ransomware, data-theft attacks

A cybercrime group is attacking U.S. organizations with ransomware for financial gain, while separately stealing data on behalf of Iran’s government from U.S., Israel, United Arab Emirates and Azerbaijan organizations.

That’s the warning the U.S. Cybersecurity and Infrastructure Security Agency (CISA) and the FBI issued this week in their joint advisory “Iran-based Cyber Actors Enabling Ransomware Attacks on US Organizations.”

The advisory outlines the Iran-based group’s tactics, techniques and procedures, as well as indicators of compromise, and provides mitigation recommendations.

 

 

The group, known as Pioneer Kitten, Fox Kitten, UNC757, Parisite, RUBIDIUM and Lemon Sandstorm, has been conducting a “high volume of network breach attempts against U.S. organizations since 2017." It’s also known as Br0k3r and xplfinder, and uses the Iranian company name Danesh Novin Sahand as a cover.

For a deep dive and analysis of the advisory, check out the Tenable blog “AA24-241A: Joint Cybersecurity Advisory on Iran-based Cyber Actors Targeting US Organizations.”

To get more details, read:

• CISA’s announcement
• The full advisory “Iran-based Cyber Actors Enabling Ransomware Attacks on US Organizations”

4 - White House requires AI inventories from federal agencies

Has your organization considered the importance of inventorying AI assets? Need guidance on how to best document your organization’s usage of AI tools? Check out how the White House is approaching this issue.

In its recently finalized “Guidance for 2024 Agency Artificial Intelligence Reporting Per EO 14110,” the White House outlines guidelines for federal agencies to compile and submit inventories of their AI use cases.

The document states that agencies must conduct an annual inventory and metrics of their AI use cases. It also details the “criteria, format, and mechanisms” agencies should use to create their reports.

For all AI use cases, agencies will have to state their purpose, expected benefits, outputs, stage of development, and whether they impact rights or safety, among other information.

For some AI use cases, the White House wants to know more details, including:

• Does it disseminate information to the public?
• Does it involve personally identifiable information?
• Has the agency’s privacy chief assessed the use case’s privacy risks?
• What agency-owned data was used to train or fine-tune the AI model in question?
• Does it include custom-developed code, and does the agency have access to the code?
• Can it make a decision or trigger an action without direct human involvement that could impact on rights or safety?

To get more details, check out the document “Guidance for 2024 Agency Artificial Intelligence Reporting Per EO 14110.”

For more information about the importance of inventorying AI assets to prevent “shadow AI” problems:

• “Do You Think You Have No AI Exposures? Think Again” (Tenable)
• “10 ways to prevent shadow AI disaster” (CIO)
• “First Step in Securing AI/ML Tools Is Locating Them” (Dark Reading)
• “Shadow AI poses new generation of threats to enterprise IT” (TechTarget)
• “The risks of shadow AI and what leaders can do to prevent it” (IT Pro)

5 - Google: APT29 exploiting known bugs in iOS and Chrome

APT29, a nation-state attacker backed by the Russian government, is actively trying to exploit known vulnerabilities in iOS WebKit and ChromeOS using techniques employed by spyware actors.

That’s according to the Google Threat Analysis Group (TAG), which this week said that over the past nine months, multiple exploit campaigns were delivered via a watering hole attack on Mongolian government websites. 

Google TAG believes “with moderate confidence” that the attacks were carried out by APT29, also known as Cozy Bear, Nobelium and Midnight Blizzard. The group is known for high-profile hacks, including breaches of Microsoft and SolarWinds.

 

 

In the campaigns observed by Google TAG, APT29 first deployed an iOS WebKit exploit of CVE-2023-41993 that affects iOS versions older than 16.6.1. Later, it launched a ChromeOS exploit chain of CVE-2024-5274 and CVE-2024-4671 targeting Android users running versions M121 to M123. Patches are available for all three vulnerabilities.

“These campaigns delivered n-day exploits for which patches were available, but would still be effective against unpatched devices,” reads the Google TAG blog titled “State-backed attackers and commercial surveillance vendors repeatedly use the same exploits.”

For more information about APT29, check out these Tenable blogs:

• “CISA Says Midnight Blizzard Swiped U.S. Gov’t Emails During Microsoft Hack”
• “Poor Identity Hygiene at Root of Nation-State Attack Against Microsoft”
• “French cyber agency: Nobelium / Midnight Blizzard is spying on European governments on Russia's behalf”
• “Midnight Blizzard swiped Microsoft’s source code, broke into its internal systems”
• “CISA, NCSC issue cloud security advisory on Russia-backed APT29”

6 - NIST updates draft of its digital-identity recommendations

The U.S. National Institute of Standards and Technology (NIST) has released a new draft of its “Digital Identity Guidelines,” which, when finalized, U.S. federal agencies will be required to adopt and follow.

The latest draft details NIST’s identity management requirements, including identity proofing and authentication. It also offers best practices for improving the privacy protection and usability of digital identity tools.

NIST’s current guidelines date from 2017. The process to update them began in 2022. 

 

 

These are some of the changes and additions in the latest draft, which is open for comment through October 7, 2024:

• A set of recommended metrics for evaluating the performance of identity management tools
• Broader requirements and recommendations for fraud management
• The addition of user-controlled digital wallets to the federation model
• A new taxonomy and structure for identity-proofing controls

The 2022 version of the draft received almost 4,000 comments from 140 organizations and individuals, many focused on digital wallets and on passkeys digital credentials.

The digital identity guidelines aim to “ensure security, privacy and accessibility during the identity-proofing process for people accessing government services,” reads a NIST statement.

To get more details, check out:

• The NIST announcement “NIST Releases Second Public Draft of Digital Identity Guidelines for Final Review”
• The latest draft of the NIST “Digital Identity Guidelines”

A joint Cybersecurity Advisory highlights Iran-based cyber actor ransomware activity targeting U.S. organizations. The advisory includes CVEs exploited, alongside techniques, tactics and procedures used by the threat actors.

Background

On August 28, the Cybersecurity and Infrastructure Security Agency (CISA) published a joint Cybersecurity advisory (CSA) in coordination with The Federal Bureau of Investigation (FBI) and the Department of Defense Cyber Crime Center (DC3). The advisory highlights the recent activities of Iranian threat actors conducting ransomware operations against US organizations across several industries including local government, defense, finance, education and healthcare as well as other countries including Israel, Azerbaijan and the United Arab Emirates.

The threat actors named in the advisory go by a few monikers including Pioneer Kitten, Fox Kitten, UNC757, Parasite, RUBIDIUM and Lemon Sandstorm. These actors have been observed to be collaborating with ransomware groups including NoEscape, Ransomhouse and ALPHV (aka BlackCat) to extort their victims. The technical aspects of the advisory highlight what techniques, tactics and procedures (TTPs) the threat actors have been observed using, including indicators of compromise (IOCs). The advisory flags six specific CVEs that are leveraged by the threat actors in the initial access phase of their attacks:

CVEDescriptionCVSSv3VPRCVE-2024-3400PAN-OS: Arbitrary File Creation Leads to OS Command Injection Vulnerability in GlobalProtect1010CVE-2024-24919Check Point Security Gateway Information Disclosure Vulnerability8.68.3CVE-2019-19781Citrix Application Delivery Controller (ADC) and Gateway (formerly NetScaler ADC and Netscaler Gateway) Directory Traversal Vulnerability9.89.4CVE-2023-3519Citrix Application Delivery Controller (ADC) and Gateway (formerly NetScaler ADC and Netscaler Gateway) Unauthenticated Remote Code Execution Vulnerability9.89CVE-2022-1388F5 BIG-IP iControl REST Remote Code Execution Vulnerability9.88.4CVE-2024-21887Ivanti Connect Secure and Ivanti Policy Secure Command Injection Vulnerability9.110

*Please note: Tenable’s Vulnerability Priority Rating (VPR) scores are calculated nightly. This blog post was published on August 28 and reflects VPR at that time.

Analysis

CVE-2024-3400 is a remote code execution vulnerability that arises from a combination of two distinct bugs in PAN-OS, specifically affecting the GlobalProtect service. The first bug is related to how the GlobalProtect service handles session IDs. The service did not sufficiently validate the format of session IDs before storing them. This oversight allowed an attacker to store an empty file with a filename of their choosing, effectively setting the stage for the exploit. The second bug involves the assumption that filenames used within the system were system-generated and therefore trustworthy. This bug enabled the filenames, which were injected by the attacker in the first step, to be used as part of a command. An attacker can exploit these two bugs to execute remote shell commands without any prior authentication. Earlier this year, this vulnerability was exploited in-the-wild as a zero-day vulnerability and was tracked at the time by Palo Alto Networks Unit 42, calling the activity Operation MidnightEclipse.

CVE-2024-24919 is an information disclosure vulnerability affecting CheckPoint Security Gateway devices configured with either the remote Access VPN or Mobile Access Software Blades enabled. This vulnerability allows an unauthenticated remote attacker to read the contents of arbitrary files located on the affected appliance. This could allow an attacker to read sensitive files. An example of one such attack path is accessing the ‘/etc/shadow’ file, which could result in the extraction of password hashes for local accounts that could potentially be decrypted. This vulnerability was also exploited in-the-wild as a zero-day, around the same time security researchers also published a proof-of-concept (PoC).

CVE-2019-19781 is a directory traversal vulnerability in Citrix Application Delivery Controller (ADC) and Gateway, formerly known as NetScaler ADC and Netscaler Gateway. This vulnerability allows an attacker to send a specially crafted HTTP request that exploits the path traversal issue, enabling unauthorized access to restricted directories on the device. Through this access, the attacker can execute arbitrary code without any need for authentication, potentially resulting in remote code execution (RCE) on the affected device. This vulnerability has been widely abused by multiple threat actors over the years and has been featured prominently in our 2020, 2021 and 2022 Threat Landscape Reports. Additionally, it has been featured in multiple blogs from Tenable Research and has been included in multiple CSA’s from CISA and other government entities across the globe.

CVE-2023-3519 is a critical RCE vulnerability in Citrix ADC and Citrix Gateway that allows an unauthenticated attacker to execute arbitrary code on the vulnerable appliances. The attack can be performed over the network, making it particularly dangerous in environments where these devices are exposed to the internet. The vulnerability stems from improper handling of specific request data, leading to memory corruption that can be exploited to gain control of the system.

CVE-2022-1388 is an iControl REST RCE vulnerability in F5 BIG-IP devices stemming from an authentication bypass bug. The flaw resides in the iControl REST interface, where improper access control allows unauthenticated users to execute arbitrary system commands with root privileges. This vulnerability is particularly dangerous because it does not require user interaction or authentication, making it easy for attackers to exploit. Successful exploitation of CVE-2022-1388 can lead to complete system compromise, enabling attackers to take full control of the device, modify configurations, exfiltrate sensitive data and use the compromised device as a launching point for further attacks within the network.

CVE-2024-21887 is a command injection vulnerability in Ivanti Connect Secure and Ivanti Policy Secure that allows remote attackers to execute arbitrary commands on the affected devices. This vulnerability occurs due to insufficient input validation in the administrative interface, which can be exploited by sending specially crafted HTTP requests. Successful exploitation can result in full system compromise, providing the attacker with the ability to execute commands with the highest privileges, potentially leading to data loss, system disruption or further propagation of malicious activity.

Legacy Vulnerabilities Remain a Looming Threat

An analysis of metadata performed by Tenable Research provides us with unique insight to two of these legacy CVEs, CVE-2019-19781 and CVE-2022-1388. From our research only about half of impacted assets have been successfully remediated. Legacy vulnerabilities present a significant risk, as threat actors frequently exploit unpatched vulnerabilities, particularly in SSL VPNs. This trend has been consistently highlighted by the Tenable Security Response Team (SRT) in their annual Threat Landscape Reports as mentioned in the section for CVE-2019-19781. To mitigate these risks, it is imperative to prioritize the remediation of legacy vulnerabilities alongside newer threats, ensuring a more comprehensive and robust security posture.

Source: Tenable Research

Ten of Thousands of Internet Facing Instances May Be Affected

It’s not surprising that threat actors are leveraging these vulnerabilities for initial access given that there are tens of thousands of potentially vulnerable devices for each of the relevant technologies discoverable on Shodan.io.

Source: Shodan.io

The results in the image above are based query results at the time this blog was composed and were obtained using the queries in the table below:

TechnologyDetection LogicQuery LinkPalo Alto Networks PAN-OSSearches for any PAN-OS instances.QueryF5 BIG-IPThe presence of "BIG-IP®- Redirect" in the title likely indicates a redirection page typically used in login portals or other access control scenarios managed by a BIG-IP device.QueryCitrix Application Delivery Controller (ADC) and GatewaySearches for favicon hash values for Citrix ADC, Gateway, AAA and VPN.QueryCheck Point Security GatewayQuery looks for servers with "Check Point SVN Foundation". This is intrinsically linked to Check Point Security Gateway devices, especially those configured with the Remote Access VPN or Mobile Access Software Blades.QueryIvanti Connect Secure and Ivanti Policy SecureQuery looks for a CGI script named "welcome.cgi" that is used to display a logo page component on the welcome or login page used by Ivanti / Pulse Secure.QuerySolution

Each of the vulnerabilities described in the CSA have been around for a period of time and each of the vendors have released the respective patches and mitigations. We recommend reviewing each of the vendors advisories shown below:

• Palo Alto Networks Security CVE-2024-3400 Advisory
• Check Point 2024-2024-24919 Advisory
• Citrix CVE-2019-19781 Advisory
• Citrix CVE-2023-3519 Advisory
• F5 CVE-2022-1388 Advisory
• Ivanti CVE-2024-21887 Advisory

Additionally, the CSA provides IoCs and technical details that may aid organizations in their incident response processes. We highly recommend reviewing the details outlined in the CSA. If your organization has assets that have not been patched for the CVE’s listed above, it’s possible that unpatched devices have been impacted due to the severity and frequency of attacks involving these vulnerabilities. As such, careful review of these systems and incident response processes may be needed to determine impact and scope of a potential compromise of unpatched systems.

Identifying affected systems

Tenable offers several solutions to help identify potential exposures and attack paths as well as identifying systems vulnerable to the CVEs mentioned in the CSA. For a holistic approach, we recommend Tenable One. The Tenable One Exposure Management Platform extends beyond traditional vulnerability management, which concentrates on the discovery and remediation of publicly disclosed CVEs. A foundational part of any exposure management program, Tenable One includes data about configuration issues, vulnerabilities and attack paths across a spectrum of assets and technologies — including identity solutions (e.g., Active Directory); cloud configurations and deployments; and web applications.

Tenable Plugin Coverage

A list of Tenable plugins for these vulnerabilities can be found on the individual CVE page’s for CVE-2024-3400, CVE-2024-24919, CVE-2019-19781, CVE-2023-3519, CVE-2022-1388 and CVE-2024-21887. This link will display all available plugins for these vulnerabilities, including upcoming plugins in our Plugins Pipeline.

Detection of legitimate tools used by adversaries and mentioned in the CSA:

ToolDetection Plugin IDAnyDesk

Plugin ID 189953 - AnyDesk Installed (Windows)

Plugin ID 189955 - AnyDesk Installed (macOS)

Plugin ID 189973 - AnyDesk Installed (Linux)

Tenable Attack Path Techniques

MITRE ATT&CK IDDescriptionTenable Attack Path TechniquesT1012Query RegistryT1012_WindowsT1059.001Command and Scripting Interpreter: PowerShellT1059.001_WindowsT1078.002Valid Accounts: Domain AccountsT1078.002_WindowsT1078.003Valid Accounts: Local AccountsT1078.003_WindowsT1098Account Manipulation: Additional Cloud Credentials/Roles

T1098.001_AWS

T1098.001_Azure

T1098.003_AWS

T1098.003_Azure

T1133External Remote Services

T1133_AWS

T1133_Azure

T1133_Windows

T1053Scheduled Task/Job: Scheduled TaskT1053.005_WindowsT1219Remote Access SoftwareT1219_WindowsT1482Domain Trust DiscoveryT1482_Windows

Tenable Identity Exposure Indicators of Exposure and Indicators of Attack

MITRE ATT&CK IDDescriptionIndicatorsT1078Dormant AccountsC-SLEEPING-ACCOUNTST1078Account with Possible Empty PasswordC-PASSWORD-NOT-REQUIREDT1078User Account Using Old PasswordC-USER-PASSWORDT1078Last Change of the Microsoft Entra SSO Account PasswordC-AAD-SSO-PASSWORDT1078AdminCount Attribute Set on Standard UsersC-ADMINCOUNT-ACCOUNT-PROPST1078Reversible Passwords in GPOC-REVER-PWD-GPOT1078Potential Clear-Text PasswordC-CLEARTEXT-PASSWORDT1078User Primary GroupC-DANG-PRIMGROUPIDT1078Domain Controllers Managed by Illegitimate UsersC-DC-ACCESS-CONSISTENCYT1078Accounts With Never Expiring PasswordsC-PASSWORD-DONT-EXPIRET1078Kerberos Configuration on User AccountC-KERBEROS-CONFIG-ACCOUNTT1078Privileged Authentication Silo ConfigurationC-AUTH-SILOT1078ADCS Dangerous MisconfigurationsC-PKI-DANG-ACCESST1078Last Password Change on KRBTGT accountC-KRBTGT-PASSWORDT1078Dangerous Sensitive PrivilegesC-DANGEROUS-SENSITIVE-PRIVILEGEST1078Logon Restrictions for Privileged UsersC-ADMIN-RESTRICT-AUTHT1078Native Administrative Group MembersC-NATIVE-ADM-GROUP-MEMBERST1078Privileged Accounts Running Kerberos ServicesC-PRIV-ACCOUNTS-SPNT1078Application of Weak Password Policies on UsersC-PASSWORD-POLICYT1078Detection of Password WeaknessesC-PASSWORD-HASHES-ANALYSIST1078Recent Use of the Default Administrator AccountC-ADM-ACC-USAGET1078Domain with Unsafe Backward-Compatibility ConfigurationC-DSHEURISTICST1098Dangerous Rights in the AD SchemaC-ABNORMAL-ENTRIES-IN-SCHEMAT1098Mapped Certificates on AccountsC-SENSITIVE-CERTIFICATES-ON-USERT1098Vulnerable Credential Roaming Related AttributesC-CREDENTIAL-ROAMINGT1098Ensure SDProp ConsistencyC-SDPROP-CONSISTENCYT1098Verify Permissions Related to Microsoft Entra Connect AccountsC-AAD-CONNECTT1098User Primary GroupC-DANG-PRIMGROUPIDT1098Domain Controllers Managed by Illegitimate UsersC-DC-ACCESS-CONSISTENCYT1098Shadow CredentialsC-SHADOW-CREDENTIALST1098Missing MFA for Non-Privileged AccountMISSING-MFA-FOR-NON-PRIVILEGED-ACCOUNTT1098First-Party Service Principal With CredentialsFIRST-PARTY-SERVICE-PRINCIPAL-WITH-CREDENTIALST1098Missing MFA for Privileged AccountMISSING-MFA-FOR-PRIVILEGED-ACCOUNT

Tenable Web App Scanning

MITRE ATT&CK IDDescriptionIndicatorsT1190Exploit Public-Facing ApplicationT1190_WASCVEDescriptionPlugin IDCVE-2024-3400Palo Alto PAN-OS GlobalProtect Remote Code Execution114282CVE-2024-24919Check Point Quantum Gateway Directory Traversal114291CVE-2024-21887Ivanti Connect Secure 9.x / 22.x Authentication Bypass114165Get more information

• Joint Cybersecurity Advisory: AA24-241A: Iran-based Cyber Actors Enabling Ransomware Attacks on US Organizations
• Tenable Blog: CVE-2024-3400: Zero-Day Vulnerability in Palo Alto Networks PAN-OS GlobalProtect Gateway Exploited in the Wild
• Tenable Blog: CVE-2024-24919: Check Point Security Gateway Information Disclosure Zero-Day Exploited in the Wild
• Tenable Blog: CVE-2019-19781: Critical Vulnerability in Citrix ADC and Gateway Sees Active Exploitation While Patches are Still Not Available
• Tenable Blog: CVE-2019-19781: Exploit Scripts for Remote Code Execution Vulnerability in Citrix ADC and Gateway Available
• Tenable Blog: CVE-2023-3519: Critical RCE in Netscaler ADC (Citrix ADC) and Netscaler Gateway (Citrix Gateway)
• Tenable Blog: CVE-2022-1388: Authentication Bypass in F5 BIG-IP
• Tenable Blog: CVE-2023-46805, CVE-2024-21887: Zero-Day Vulnerabilities Exploited in Ivanti Connect Secure and Policy Secure Gateways
• Tenable Blog: CVE-2023-46805, CVE-2024-21887, CVE-2024-21888 and CVE-2024-21893: Frequently Asked Questions for Vulnerabilities in Ivanti Connect Secure and Policy Secure Gateways
• Tenable Blog: CISAs 2022 Top Routinely Exploited Vulnerabilities (AA23-215A)
• Tenable’s 2020 Threat Landscape Retrospective
• Tenable’s 2021 Threat Landscape Retrospective
• Tenable’s 2022 Threat Landscape Report

Join Tenable's Security Response Team on the Tenable Community.

Learn more about Tenable One, the Exposure Management Platform for the modern attack surface.

DSPM solutions provide a comprehensive, up-to-date view into cloud-based data and risk. An integrated CNAPP and DSPM solution elevates this analysis to expose toxic combinations and security gaps across cloud environments.

As organizations ramp up their use of cloud-native applications, the amount of sensitive data stored in the cloud grows – as does the difficulty in managing and scaling data-related risk management and compliance. Hackers are motivated to get at data stored in the cloud. Employee data, customer information, business IP – it’s all (un)fair game. 

Enter cloud native application protection (CNAPP) solutions. In light of the massive increase in data-related breaches - and their cost, integrating data security posture management (DSPM) in CNAPP is essential to reduce risk. It also simplifies security efforts, improves compliance and ensures that data security is an integral part of your overall security strategy.

See the demo

The superpower of a unified CNAPP

A quick concept review.

A CNAPP offers security and compliance for cloud-native applications throughout their lifecycle, across multiple clouds. Its superpower lies in providing a unified view and contextual analysis across infrastructure, workloads, identities and more. A CNAPP isolates exposures, including hidden toxic combinations, and pinpoints how to fix them. 

DSPM tools – a $94 billion market projected to double by 2031(InsightScan) – focus specifically on security and compliance-readiness for data in the cloud. These tools continuously scan the environment to find data, including databases, object storage and data lakes, across cloud and service providers, flowing to or from, any location. They classify and protect data assets, ensure audit security policies and requirements, and detect data-related threats. 

Using DSPM alone makes it difficult to maintain a centralized and deep view into sensitive data; where it’s stored, what kind it is, who can access it and how it is used. It’s like getting a view into the stars but not the galaxy. Inside a CNAPP, DSPM gives the needed illumination and context.

Let’s explore how a DSPM works.

Comprehensive visibility into data assets

A key DSPM function is to continuously provide a comprehensive, up-to-date view into cloud based data assets and risk. When joined with cloud security posture, this data analysis exposes discreet security gaps and toxic combinations and, importantly, the impact on data if exploited. Teams gain greater prioritization accuracy and focus around findings, mitigating alert fatigue. 

Figure - Integrated DSPM in a CNAPP enables powerful permission queries into specific types of sensitive data, such as digital identities, for prioritized focus on risk exposure Better data security and compliance 

By implementing a DSPM-integrated CNAPP, you can reduce the risk of data breaches and non-compliance. DSPM solutions continuously assess compliance posture and ensure that cloud data is classified, protected and accessed according to policies and frameworks such as GDPR, HIPAA, and CCPA. You can automate policy enforcement and generate comprehensive, audit-ready reports to reduce compliance fire drills. Users can act quickly upon discovering misconfigurations, unauthorized access and potential security threats discovered in near real-time.

Managing data risk proactively 

Proactive risk management is a cornerstone of cloud security. Integrating DSPM empowers teams to detect potential data risks, as well, and take action early, before they become big problems. Through the use of advanced analytics, DSPM capabilities detect anomalous patterns and behaviors that can indicate a security threat to your sensitive data. A DSPM integrated CNAPP provides actionable insights and recommendations for improving your overall data security posture, ensuring your organization stays a step ahead of evolving threats.

Faster, streamlined incident response

In the event of a security incident, time and action is everything. Integrated DSPM plays a vital role in streamlining the incident response process. By providing near real-time alerts and detailed forensic data, the solution helps security teams quickly identify the scope and impact of a breach. These insights speed up the process, enabling stakeholders to achieve containment and remediation and minimize potential damage. DSPM's integration with other CNAPP components, such as workload protection and cloud security posture management, enables a coordinated, efficient incident response, reducing downtime and maintaining business continuity. 

Use case: Tenable Cloud Security with integrated DSPM

Tenable Cloud Security isolates and eradicates cloud risks across infrastructure, workloads, identities and data. Having recently acquired Eureka Security, we are now integrating leading DSPM capabilities into our CNAPP context mix - stay tuned as new features roll out.

In this use case, we show how you can use powerful permission querying to detect and filter for resources with certain types of sensitive data, such as digital identity or financial information, to understand and explore your risk exposure, and focus on prioritized security findings. 

Conclusion

Integrated DSPM is an indispensable component of a robust CNAPP strategy. It extends comprehensive visibility and deep risk context to data assets, safeguarding data and keeping your security posture strong through automation and actionability.

The cloud, artificial intelligence (AI), machine learning and other technological breakthroughs are radically changing the modern work environment. New assets and services offer increased flexibility, growth potential and access to more resources. However, they also introduce new security risks. Managing vulnerabilities across this ever-expanding threat landscape requires a risk-based approach beyond point solutions and reactive patch management.

The cyberthreat landscape is rapidly evolving. Attackers are targeting your increasingly complex and expanding attack surface, knowing cyber professionals struggle to manage the daunting task of securing this vast digital environment.

A recent report from Cybersecurity Ventures paints an alarming picture of just that. This year, cybercrime damage costs will likely reach $9.5 trillion globally, an issue compounded by the increased average cost of a single data breach, which, according to IBM's Cost of a Data Breach 2024 report, is $4.88 million.

Recent high-profile breaches, like the ones that have exploited the MOVEit Transfer vulnerability (CVE-2024-5806), drive this point home. It demonstrates the urgent need for organizations of all sizes to implement robust vulnerability management practices. Yet, those who don't approach this from a risk-based perspective are likely to continue to fall behind.

Such attacks underscore the critical need for risk-focused strategies to identify, prioritize and mitigate security vulnerabilities before attackers can exploit them. This means implementing and maturing vulnerability management practices to stay ahead of business risk by adopting a holistic approach that identifies and addresses actual risk head-on.

Vulnerability management beyond the compliance checkbox

Traditionally, many organizations viewed vulnerability management as a tick-the-box compliance exercise. That approach is no longer effective.

A single successful cyberattack that exploits a critical vulnerability like the MOVEit Transfer bug can expose millions of records. That's why organizations can't continue to treat vulnerability management as a one-time “set it and forget it” task. In today's dynamic threat landscape, risk-focused vulnerability management must become an integral and continuous part of your cybersecurity strategy. Here's why:

Attack surfaces are complex. The modern attack surface is no longer confined to a data center or on-prem systems and networks. IT security teams are responsible for sprawling attack surfaces with tens of thousands — or more — of assets, each with multiple vulnerabilities of varying severity at any given point. IT assets in modern enterprises now include:

• Cloud infrastructure and services
• Web applications
• Remote workforces
• Vast internet-facing devices
• IoT and OT environments

This means modern vulnerability management must be equally comprehensive and flexible as these evolving attack surfaces.

All vulnerabilities are not the same. For actionable security controls, vulnerability prioritization based on real business risk — not on an arbitrary vulnerability scoring system — is key. Focusing resources on the most critical vulnerabilities with the greatest potential to impact your operations ensures targeted, effective resolution and maximizes your time and effectiveness.

Time is critical. Timely patching and fixing of vulnerabilities are crucial. When you have unaddressed critical vulnerabilities, they create exploit opportunities for threat actors. Efficient vulnerability management decreases this window to protect your organization from potential breaches.

Collaboration is crucial. Risk-focused vulnerability management is more than cross-department collaboration. It must also extend to your key stakeholders, vendors and partners.

Traditional silos between IT, security, business units and others impede progress. The old school of thought was security was an IT problem. The reality is security is everyone's responsibility.

With increased collaboration, you can build an organization-wide culture with a holistic understanding of security risks. And, aligning risk levels with business goals facilitates coordinated, actionable threat response, which can have meaningful impact on business goals.

6 steps to build a risk-based vulnerability management program

1. Ensure a comprehensive asset inventory

Before securing your assets, you must know what you're protecting. That includes managing systems or devices that are short-lived or changing in your environment.

A comprehensive and up-to-date asset inventory should be the cornerstone of your vulnerability management program. This involves identifying all hardware and software assets across your entire IT environment, on-prem and in the cloud. It extends into IoT and OT environments, software and app development and more.

Regular asset audits are essential for accurate inventory. But it goes beyond just knowing which assets you have. You must also have a deep understanding of:

• Where they're located
• Who uses them
• How critical each is to your operations

This knowledge, coupled with threat intelligence, empowers your teams to prioritize vulnerabilities based on asset criticality.

1. Continuously assess and monitor vulnerabilities

Threat actors don't take breaks, and neither can your vulnerability management controls. Implementing real-time vulnerability assessment and continuous scanning tools enables early detection of new vulnerabilities and emerging cyber risks. This proactive approach empowers your teams to address issues before they become major incidents like a security breach. Integrating automation also matures your cyber defenses by reducing manual tasks, decreasing time and errors, and ensuring consistent, continuous oversight.

1. Focus on what matters most

Critical or high Common Vulnerability and Exposures Scoring System vulnerabilities (CVSS scores) aren't always pressing threats. Some vulnerabilities pose more risk to your organization than others, regardless of an arbitrary risk score.

Risk prioritization enables teams to focus limited resources on vulnerabilities most likely to affect your operations. Risk-focused mitigation can improve your security posture and make your business more resilient.

1. Close security gaps before attackers exploit them

Timely and accurate patching ultimately mitigates the most risks associated with vulnerabilities. It's essential to establish structured patch-management processes and clear timelines. Assigning roles, responsibilities and accountability ensures your team can promptly address your exposures. This closes security gaps, decreases your exposure window, and decreases the likelihood of a potential data breach.

1. Integrate industry-trusted threat intelligence

Most security teams don't have the time, resources or skills to keep pace with evolving vulnerabilities. That's why it's important to use vulnerability management solutions with built-in, industry-trusted threat intelligence, like the one provided by Tenable Research. 

Vulnerability threat intelligence, automation tools and AI enhance your insight into real-world tactics attackers use. By understanding attack vectors and methods, your teams can prioritize vulnerabilities based on exploit likelihood. From there you can proactively adjust your security controls to stay one step ahead.

1. Report, measure and benchmark

Vulnerability data is key to tracking program effectiveness. A vulnerability management platform with customized and automated reports can help your teams communicate security risks in business context.

Consider using a vulnerability management tool that collects security data in real-time. This can help your teams identify trends and see security gaps before attackers do. Track and measure the impact of your vulnerability management program to refine vulnerability management processes and narrow your attack surface.

5 tips to mature your vulnerability management practices

Developing a robust vulnerability management program is just the first step in protecting your attack surface. Effective vulnerability management is an ongoing process. Here are five tips to help evolve and mature your program as the threat landscape evolves.

• Standardize and automate processes
  • Develop standardized vulnerability scanning, risk assessment, prioritization, remediation and reporting processes.
  • Use automated vulnerability assessment and management tools to streamline manual tasks, decrease errors and free up your security teams to focus on other needs.
• Invest in vulnerability management tools
  • Consider a vulnerability management solution like Tenable Vulnerability Management. It has comprehensive scanning capabilities, real-time threat intelligence and risk-based prioritization. 
  • Tenable's Vulnerability Priority Rating can help you assess risk in your unique environment. It’s not arbitrary scoring. The solution also has advanced reporting.
  • Also, consider conducting internal and external penetration testing to ensure your controls function as intended.
• Empower your security teams
  • Provide your teams with the necessary training, resources, security tools and threat intelligence to effectively utilize a vulnerability management platform. If you have limited resources or team availability, consider partnering with a security consultant like Tenable to customize your vulnerability management program.
• Promote security awareness
  • Foster a culture of security awareness. 
    • Educate employees, key stakeholders, vendors and partners about security best practices and your policies and compliance requirements.
  • Ensure your security teams have tools, security policies and procedures to manage key vulnerability management tasks such as:
    • Risk management
    • Patching vulnerabilities
    • Reporting suspicious activities
• Keep improving
  • Track and measure key performance indicators (KPIs) such as:
    • The number of vulnerabilities identified
    • Risk tolerance and threshold
    • Time to remediation
    • Risk reduction
  • Use key security metrics to find areas for improvement.
  • Close security gaps.
  • Routinely refine your vulnerability management program.

Tenable Vulnerability Management benefits

With Tenable Vulnerability Management and Tenable Security Center, you get a comprehensive view of all the assets and vulnerabilities on your network so you can understand cyber risk and know which vulnerabilities to fix first. The key difference between the solutions is how they are managed. Tenable Vulnerability Management is managed in the cloud, while Tenable Security Center is managed on-prem.

It provides:

• Comprehensive visibility
  • Use comprehensive asset discovery tools and vulnerability scanner capabilities.
• Risk-based prioritization
  • Remediate threats that pose the greatest risk to your organization.
• Automated workflows
  • Streamline and automate vulnerability management with best practice processes and tools.
  • Save time, resources and money.
  • Decrease the chance of human errors or compliance oversights.
• Actionable insights
  • Get insights from industry-leading threat intelligence.
  • Customize reporting to make more informed and business-aligned security decisions.
  • Enable continuous improvement driven by Tenable's expertise and ongoing vulnerability research to stay ahead of evolving threats

A comprehensive, risk-based vulnerability management program can help your organization significantly reduce your attack surface and mature your security posture. A solution like Tenable Vulnerability Management can help you reach these goals. It has tools and insights to proactively identify, prioritize and remediate vulnerabilities and other security weaknesses across your enterprise.

Building and maturing your vulnerability management program is no longer optional. You need risk-based vulnerability management as part of a mature security strategy. Effective vulnerability management isn't just about the technologies and resources you use. It's about adopting best practices to align your vulnerability management program with business goals to ensure business resilience.

See how Tenable Vulnerability Management can help your organization proactively protect and defend your entire attack surface.

Empowering K-12 schools and libraries to strengthen their cybersecurity posture with new funding opportunities and best practices.

Recent attacks reveal that K-12 schools and libraries are prime targets for cyber actors. Ransomware attacks have had dire consequences and highlighted the need for such institutions to strengthen their cybersecurity posture and reduce risk. Increased attacks against schools have resulted in financial losses, sharing of sensitive data and loss of learning time. Additionally, attacks against libraries have resulted in sharing sensitive data and losing vital services, with recovery taking weeks to months. 

Despite these attacks, cybersecurity is often a challenge for these institutions due to underfunding, limited financial and staffing resources, and competing priorities. This often results in schools and libraries taking a reactive vs. proactive approach to cybersecurity, with cyber resilience becoming a secondary focus. By taking advantage of the FCC's new $200 million Cybersecurity Pilot Program, these establishments can access resources to help protect themselves against these growing threats and ensure the safety of their digital environments.

What can schools and libraries do to reduce cyber risk?

The FCC recently announced cybersecurity funding for K-12 schools and libraries. The program will enable eligible schools and libraries to request funding for cybersecurity services and equipment. The application window is expected to open in the Fall of 2024. More information about the program can be found on the FCC Cybersecurity Pilot Program webpage.

The details:

• Funding Amount: $200 million in Universal Service Funds available to eligible schools and libraries for eligible cybersecurity services and equipment.
• Timeframe: Three (3) years
• Eligible Recipients: Public and non-profit private schools (K-12), libraries, and groups of schools and libraries (e.g., consortia, districts, systems) that meet the E-Rate program’s eligibility requirements. 
• Awards: Eligible schools selected for the Pilot Program may receive a minimum budget of $15,000 and a maximum budget of $1.5 million. Eligible libraries selected for the Pilot Program may receive up to $175,000. Participants may seek reimbursement for Pilot Program expenses as they are incurred, as long as the total amount of funding does not exceed the total amount awarded for the three-year term.
• Services/equipment covered: Eligible services and equipment include vulnerability management, identity protection and authentication; monitoring, detection, and response; endpoint protection; and advanced/next-generation firewalls. Here’s a detailed list of eligible services and equipment. 
• Has something similar been done before? The Pilot will operate similarly to the E-Rate program which provides discounts for telecom and internet services for schools and libraries.

Specifics on the application process 

The application form is divided into two parts. Part one will be open to all eligible schools and libraries, and applicants will be required to submit the first part of FCC Form 484, titled “Schools and Libraries Cybersecurity Pilot Program Application.” This form will be available and submitted online through the pilot portal on the USAC website once the application window opens in the fall. 

The second part of the FCC Form 484 application is only for applicants selected for the program. Information about the application process and requirements can be found on the FCC Cybersecurity Pilot Program webpage. Applicants can now begin to prepare for the application window opening up in the fall by completing pre-application steps found in the FCC program application guide.

Cybersecurity Pilot Program requirements

As required by the FCC, to apply for funding schools and libraries will need to submit a plan that shows how funding will be used to address cybersecurity needs and reduce cyber risk. The project plan should identify cybersecurity goals, outline steps to achieving them, and include evaluation metrics.

Tenable is uniquely positioned to help schools and libraries reduce cyber risk through the Tenable One Exposure Management Platform. As the only end-to-end exposure management platform, Tenable One radically unifies security visibility, insight and action across the attack surface. Using Tenable One, schools and libraries can take a proactive stance towards cybersecurity and address critical concerns, such as protecting against ransomware.

With diverse IT environments and many users in their directory, Tenable one helps schools gain full visibility into their entire infrastructure and address privilege solutions such as Active Directory. 

As schools and libraries face staffing and resource constraints, Tenable One enables them to prioritize effectively, by anticipating threats and focusing efforts to prevent likely attacks. Using insights from the platform, schools and library leaders can accurately communicate cyber risk, measure progress over time, and communicate results to stakeholders such as the board, state, or auditors.

Tenable One allows you to:

• Gain a unified view of cyber risk across your attack surface, exposing the gaps that leave you vulnerable to attack across all types of assets and pathways, spanning IT, OT and IoT assets, cloud resources, containers, web apps, identity access and privilege solutions, and more.
• Analyze cyber risk and insights from across the attack surface as one, connecting dots to identify the true exposures threatening your school or library, reputation and trust. 
• Communicate risk posture to stakeholders in order to make informed decisions and mobilize resources to find and fix exposures with the highest likelihood of attack and disruption to your school or library.

 

Learn more:

• Tenable One website
• Tenable One datasheet
• Tenable Vulnerability Management product page

Conventional wisdom suggests the keys to protect critical infrastructure against cyberattacks are network segmentation and OT security. But continued breaches imply those methods alone fall short. In fact, a CISA probe of 121 critical infrastructure networks found that their weakest link is identity compromise. Learn how to leverage an attacker’s perspective to better secure critical infrastructure.

The threat landscape for critical infrastructure has become increasingly perilous, fueled by nation-state attackers and other well-organized threat actors. Spanning 16 sectors vital to U.S. national security, including manufacturing, financial services and government, critical infrastructure organizations face relentless cyberattacks designed to disrupt services, exfiltrate sensitive information or hold entire operations for ransom. 

Because compromising operational technology (OT) is frequently a key objective, security teams understandably focus heavily on OT security and network segmentation. But this strategy has proven too narrow, leaving out critical considerations, such as the role of identity in breaches. In fact, identity compromise offers attackers the most significant exposure gap to gain initial access to critical infrastructure networks and to move laterally.

In this blog, we explore high-profile breaches involving OT and critical infrastructure; look at the role that identity and other MITRE ATT&CK techniques played in those attacks; and provide practical recommendations to improve the security of critical infrastructure by integrating security insights from multiple security domains. 

Identity compromise – a security weak link in critical infrastructure 

The Cybersecurity and Infrastructure Security Agency (CISA), along with the U.S. Coast Guard (USCG), probed the networks of 121 critical infrastructure organizations. Formally called Risk and Vulnerability Assessments (RVAs), these probes were intended to assess an organization’s network capabilities and network defenses against potential threats. The probes map the results to the MITRE ATT&CK framework.

A mid-2023 CISA report and accompanying charts detailed the findings of these RVAs, including that 90% of the time, initial access was gained via identity compromise. The report also said that when combined with other techniques, identity also represents the primary means of privilege escalation, and in turn, an enabler of lateral movement between IT and OT environments. 

This finding is eye-opening, given the widespread use of air gapping, network segmentation and substantial investments in IT and OT security.

MITRE ATT&CK techniques used to gain initial access to critical infrastructure networks

(Source: “CISA Analysis: Fiscal Year 2022 Risk and Vulnerability Assessments” report, June 2023)

The window for disclosing breaches is now 72 hours

Compounding the urgency of the elevated threat landscape, security teams face heightened pressure for transparency and accountability from new legislation aimed at critical infrastructure industries across many countries. The U.S. Cyber Incident Reporting for Critical Infrastructure Act of 2022 (CIRCIA), and the Network and Information Systems (NIS2) Directive in the European Union, for example, mandate that cyber incidents having material impact on critical infrastructure must be disclosed within 72 hours. Beyond increased visibility and accountability, these mandates dramatically shrink the window from the time a suspected breach is identified to when it must be disclosed, driving the need for more effective, preventative security. 

Worth noting, similar mandates have also been enacted for other industries beyond the scope of critical infrastructure.

Network segmentation and OT security alone are insufficient

When it comes to attack execution, the reality is there is much overlap between how attackers gain an initial foothold, escalate privileges and move laterally. This is because fundamentally, all actions that an attacker executes leverage the privileges of either a human or a machine identity. What differs is which MITRE ATT&CK techniques are used to gain those privileges. 

Let’s look at three examples that are highly representative of critical infrastructure breaches:

Colonial Pipeline: The largest publicly disclosed cyberattack against critical infrastructure in the U.S. began when attackers gained initial access to the IT network through a compromised user password and VPN account. Remote desktop protocol was then used to deploy ransomware which exploited unpatched vulnerabilities and escalated privileges. The ransomware then encrypted critical systems. Although OT systems were not directly compromised by the attack, security teams were forced to shut down the oil pipeline to prevent the ransomware from spreading to the OT network. 

Techniques used in the Colonial Pipeline breach

Oldsmar: When attackers targeted the water treatment facility for the town of Oldsmar, Florida, it is believed that they gained an initial foothold using compromised user credentials. Once inside, they took control over an operator’s desktop using existing remote desktop software to access plant control systems, and executed commands which changed sodium hydroxide levels in the water supply to dangerously high levels. Luckily, the operator witnessed the breach and was able to intervene in time.

Techniques used in the Oldsmar breach

NotPetya: One of the most significant and destructive cyberattacks in recent history, NotPetya began with compromised credentials which allowed attackers to gain access to popular accounting software and modify its code. The software provided attackers a backdoor to distribute malware into target organizations. Once deployed, the malware spread quickly using EternalBlue, an exploit which targets a known SMP vulnerability. The Mimikatz post-exploitation tool was used to harvest credentials and propagate laterally often using legitimate Windows administrative and network sharing tools. The breach disrupted critical infrastructure across the globe, from ports run by shipping giant Maersk to radiation-monitoring systems in Chernobyl’s nuclear power plant. 

Techniques used in the NotPetya breach

In none of these cases were OT assets the initial target. But they all involved some form of human- or machine-identity compromise, and spread using a combination of existing vulnerabilities and legitimate remote-management vehicles to escalate privileges, move laterally and achieve a desired goal. 

MITRE ATT&CK techniques used to move laterally in critical infrastructure networks

(Source: “CISA Analysis: Fiscal Year 2022 Risk and Vulnerability Assessments” report, June 2023)

You can’t secure critical infrastructure without complete context

The challenge with virtually every point security tool is that they lack a comprehensive and integrated view of asset, identity and risk relationships across the attack surface. For example, OT security tools frequently lack visibility into IT assets and identities within the OT environment, including weaknesses that can be exploited to ultimately compromise critical infrastructure systems. It is these cross-domain relationships that enable initial access, lateral movement and privilege escalation. Without them, there is no way to effectively distinguish a sea of siloed alerts from true exposures which can disrupt the integrity and continuity of critical infrastructure. 

The Tenable One Exposure Management Platform is uniquely able to address these challenges. It provides complete visibility across the modern attack surface, bridging security silos and prioritizing the detection and closure of exposures that can have a material impact on critical infrastructure. Tenable One integrates rich configuration data and risk data on OT and IT assets from Tenable OT Security, with privilege and risk data on human and machine identities from Tenable Identity Exposure. With it, security staff can visualize and remediate attack paths that traverse converged IT and OT environments. 

This short video demonstrates how Tenable One solves the critical infrastructure exposure problem – disrupting attack paths before attackers can exploit them.

By adopting a horizontal approach to security that emphasizes comprehensive visibility, contextual risk assessment and prioritization of true exposure, organizations can enhance their security posture, drive efficiencies and better protect critical infrastructure from evolving threats. 

To learn more about exposure management, download the whitepaper “Hackers Don’t Honor Security Silos: 5 Steps To Prioritize True Business Exposure.”

Looking to sharpen your team’s event logging and threat detection? A new guide offers plenty of best practices. Plus, the FAA wants airplanes to be more resilient to cyberattacks. Meanwhile, check out the critical vulnerabilities Tenable discovered in two Microsoft AI products. And get the latest on ransomware trends, vulnerability management practices and election security!

Dive into six things that are top of mind for the week ending August 23.

1 - Guide outlines logging and threat detection best practices

As attackers double down on the use of stealthy, hard-to-detect “living off the land” (LOTL) techniques, cybersecurity teams should improve how they log events and detect threats. That’s the call government agencies from multiple countries made in a joint guide published this week and titled “Best Practices for Event Logging and Threat Detection.” 

“Event logging supports the continued delivery of operations and improves the security and resilience of critical systems by enabling network visibility,” reads the guide, which was developed by the Australian Cyber Security Centre (ACSC).

 

The guide, whose intended audience includes senior IT and operational technology (OT) leaders and operators, network administrators and critical infrastructure providers, groups its best practices under four categories:

• Enterprise-approved event logging policy, which includes event-log quality and the consistency of content, formats and timestamps
• Centralized log collection and correlation, which focuses on logging priorities for enterprise networks, OT systems, mobile devices and cloud environments
• Secure storage and event-log integrity, which touches on protecting logs from unauthorized access, modification and deletion
• Detection strategy for relevant threats, which deals with LOTL techniques, such as attackers’ use of legitimate tools and capabilities in the breached environment

The guide’s recommendations can help “detect malicious activity, behavioral anomalies, and compromised networks, devices, or accounts,” reads a statement from the U.S. Cybersecurity and Infrastructure Security Agency (CISA), which collaborated with the ACSC, along with cyber agencies from seven other countries.

To get more details, check out:

• CISA’s announcement
• The full 17-page guide “Best Practices for Event Logging and Threat Detection”

For more information about event logging and threat detection:

• “Security log management and logging best practices” (TechTarget)
• “Logging Cheat Sheet” (OWASP)
• “Network security logging and monitoring” (Canadian Centre for Cyber Security)
• “Introduction to logging for security purposes” (U.K. NCSC)

2 - FAA wants to tighten aviation cyber rules

To beef up the aviation sector’s cyber defenses, the U.S. government this week proposed new cybersecurity rules for airplanes, engines and propellers as they become increasingly connected to computer networks and services.

In a proposed rulemaking notice, the U.S. Federal Aviation Administration (FAA) said some of its regulations are “inadequate and inappropriate to address the cybersecurity vulnerabilities caused by increased interconnectivity.”

Current designs for airplanes can make them vulnerable to cyber risks from maintenance laptops; airport or airline networks; the internet; wireless sensors and their networks; USB devices; and cellular and satellite systems and communications; and more.

 

 

The proposed new and revised rules seek to protect airplanes, engines and propellers from “intentional unauthorized electronic interactions” (IUEI) so the FAA wants to require product designers and makers to “identify” and “assess” IUEI risks and mitigate them.

To that end, they would need to conduct a security risk analysis to identify all cyberthreats, assess threat severity, determine exploitation likelihood and mitigate these security issues.

The proposed rules are now open for public comment.

To get more details, check out the 36-page notice of proposed rulemaking titled “Equipment, Systems, and Network Information Security Protection.”

For more information about aviation cybersecurity:

• “TSA, FAA Requirements Emphasize Cybersecurity for Airport and Aircraft Operators and Airport Terminal Projects” (Tenable)
• “Aviation Cybersecurity: Risks and Mitigations” (National Business Aviation Association)
• “Why aviation needs to prioritise cybersecurity” (Airport World)
• “Protecting The Aviation Sector From Cyberattacks” (Tenable)

3 - Tenable finds critical vulnerabilities in two Microsoft AI products

Tenable Research recently discovered critical vulnerabilities in Microsoft’s Azure Health Bot Service and Copilot Studio.

In the case of Azure Health Bot Service, a cloud platform for deploying AI-powered virtual health assistants, the critical vulnerabilities allowed researchers access to cross-tenant resources within this service. 

Meanwhile, a server-side request forgery (SSRF) vulnerability in Copilot Studio allowed researchers access to potentially sensitive information regarding service internals with potential cross-tenant impact. With Copilot Studio, you can build custom Copilot conversational applications for performing large language model (LLM) and generative AI tasks.

 

 

To get all the details, read these Tenable blogs:

• “SSRFing the Web with the Help of Copilot Studio”
• “Compromising Microsoft's AI Healthcare Chatbot Service”

You can also find media coverage of the two discoveries here:

Copilot Studio

• “Microsoft Copilot Studio Exploit Leaks Sensitive Cloud Data” (Dark Reading)
• “Microsoft Copilot Studio Vulnerability Led to Information Disclosure” (SecurityWeek)
• “Microsoft Patches Critical Copilot Studio Vulnerability Exposing Sensitive Data” (The Hacker News)

Azure Health Bot Service

• “Microsoft's AI Health Bot required patching for privilege vulnerability” (Healthcare IT News)
• “Fix for Azure Health Bot vulnerabilities prevents exploitation” (TechTarget)
• “Microsoft Azure AI Health Bot Infected With Critical Vulnerabilities” (Dark Reading)

4 - Tenable polls webinar attendees on VM practices

During our recent webinar “From Frustration to Efficiency: Optimize Your Vuln Management Workflows and Security with Tenable,” we polled attendees on their biggest VM challenge and on patching prioritization. Check out what they said!

(231 webinar attendees polled by Tenable, August 2024)

(234 webinar attendees polled by Tenable, August 2024)

Want to learn how to improve key vulnerability management practices, including remediation prioritization? Watch this on-demand webinar “From Frustration to Efficiency: Optimize Your Vuln Management Workflows and Security with Tenable.”

5 - Report: Ransomware attacks jumped in July, as attackers turn to infostealer malware

Ransomware attacks spiked 20% globally in July, compared with June, as the RansomHub gang emerged as the most active group.

However, ransomware attacks were down compared with July 2023, and “remain much lower” compared with the activity observed between February to May of this year.

That’s according to the “Monthly Threat Pulse” report for July 2024, published this week by NCC Group’s Global Threat Intelligence team.

“Whether this increase reflects the start of an upward trend remains to be seen,” the report reads.

Global Ransomware Attacks by Month 2023 - 2024

(Source: “Monthly Threat Pulse” report from NCC Group, August 2024)

The industrials sector was the hardest hit, receiving about a third of all attacks, a clear sign of ransomware groups’ strong interest in breaching critical-infrastructure organizations, the report says.

Ransomware groups pounced on CVE-2024-37085, an authentication-bypass vulnerability in the VMware ESXi hypervisor product, a reminder that organizations need to continue to prioritize patching high-risk bugs.

While vulnerability exploitation remains a popular tactic for ransomware attackers, they’re also increasing their use of information stealer malware, which offers them a “far easier, faster and often cheaper” way to access a network via the use of stolen credentials, the report reads.

“The rise in sophisticated techniques, such as the use of information stealer malware in their pre-attack phase, highlights that cybercriminals are not standing still. As these threats evolve, so must our defences,” Ian Usher, Deputy Head of Threat Intelligence at NCC Group, said in a statement.

To get more details, check out:

• The report’s announcement
• The “Monthly Threat Pulse Review of July 2024”

For more information about ransomware trends:

• “Ransomware Is ‘More Brutal’ Than Ever in 2024” (Wired)
• “Ransomware on track for record profits, even as fewer victims pay” (SC Magazine)
• “Ransom recovery costs reach $2.73 million” (Help Net Security)
• “Ransomware report finds 43% of data unrecoverable after attack” (SC Magazine)

6 - CISA: Ransomware won’t impact U.S. presidential election

Although ransomware gangs may try to disrupt the U.S. general election, their attacks won’t compromise the voting and counting processes, according to CISA and the FBI. At best, ransomware attacks would cause isolated delays and be minimally disruptive.

“The public should be aware that ransomware is extremely unlikely to affect the integrity of voting systems or the electoral process,” FBI Cyber Division Deputy Assistant Director Cynthia Kaiser said in a statement.

The reason: U.S. election officials have put in place what CISA and the FBI call a “multi-layer approach to security” with a variety of technical, physical and procedural safeguards.

“In the event of a ransomware event affecting their offices, election officials have plans and redundancies in place to allow voting operations to continue so that all eligible voters are able to cast their ballot securely,” reads a joint public-service announcement from CISA and the FBI.

To get more details, check out:

• CISA’s announcement “Ransomware Disruptions During Voting Periods Will Not Impact the Security and Resilience of Vote Casting or Counting”
• The joint PSA 

Tenable Research discovered a critical information-disclosure vulnerability in Microsoft’s Copilot Studio via a server-side request forgery (SSRF), which allowed researchers access to potentially sensitive information regarding service internals with potential cross-tenant impact.

Introduction

In this blog, we take a look at a server-side request forgery (SSRF) vulnerability in Copilot Studio that leveraged Copilot’s ability to make external web requests. Combined with a useful SSRF protection bypass, we used this flaw to get access to Microsoft’s internal infrastructure for Copilot Studio, including the Instance Metadata Service (IMDS) and internal Cosmos DB instances. Before we jump into that, we’ll cover a bit of background.

 

 

Background

Recently we were looking into a couple of SSRF vulnerabilities in the APIs for Azure AI Studio and Azure ML Studio, which happened to be patched before we could report them. Thus, we looked into another area of the studios and determined that there was a bypass for SSRF protections in a separate, but similar API, which we then reported to Microsoft’s Security Response Center (MSRC). 

An SSRF vulnerability occurs when an attacker is able to influence the application into making server-side HTTP requests to unexpected targets or in an unexpected way. 

For instance, a common feature in many modern applications which deal with data analysis or machine learning is to integrate data from external services. In order to do so, the application needs to make HTTP requests to connect to those external service APIs. If an attacker is able to control the target of those requests, they could point the request to a sensitive internal resource for which the server-side application has access, even if the attacker doesn’t, revealing potentially sensitive information.

In the context of cloud applications, a common target is the IMDS, which, depending on the cloud platform, can yield useful, potentially sensitive information for an attacker (Azure, AWS, and Google all have their own versions of this), localhost, or other internal infrastructure. For this reason, many features which could potentially lead to SSRF vulnerabilities will block the feature from targeting IMDS and non-routable IPs, but where there are restrictions, there are bypasses. 

In the case of Azure AI Studio and Azure ML Studio, while the feature we attacked would block requests targeting localhost/127.0.0.1, we could point the request to an attacker-controlled server, which would then use an HTTP 301 (or 302) redirect to redirect the server’s request back to localhost/127.0.0.1. In this way, we were able to bypass their restriction, and obtain sensitive information, such as other hosts and endpoints on the network, from 127.0.0.1:4191 (a Linkerd related metrics endpoint).

With that background out of the way, let’s take a look at Microsoft’s Copilot Studio. 

Copilot Studio HttpRequestActions

Microsoft’s Copilot Studio is built on top of its Power Platform, and as a result is a pretty cool product which lets you build custom Copilots – conversational applications which let you perform a wide variety of large language model (LLM) and generative AI tasks leveraging data ingested from your Microsoft 365 environment or any other data your Power Platform environment has access to.

When creating a new Copilot, you are able to define Topics, which let you specify key phrases which a user could say to your Copilot that could lead to the AI responding in specific ways, or taking certain actions. When first exploring this feature we noticed that one of the actions that can be performed, when triggered by a key phrase, is an HTTP request.

 

 

This was immediately exciting. Any time a cloud service presents the option to perform HTTP requests on your behalf there is some testing to be done. Better yet, this HttpRequestAction (so named in the topic code editor view) allows control over HTTP request headers, which will come in handy for testing against the IMDS, since it requires special request headers.

For an initial test, and to get an idea of how the request made by the Copilot looks, we can add a header to the HttpRequestAction and make a request to our Burp Collaborator instance. 

 

 

Burp Collaborator soon shows a request made from an Azure IP address which includes our test header, and a few headers we did not add which are associated with Azure services.

Hitting IMDS

Now that we know we can make requests, we can try requesting some interesting cloud resources like the IMDS we mentioned earlier. Unfortunately, pointing the request to the url http://169.254.169.254/metadata/instance?api-version=2021-02-01 and sending the request yields a System Error response. 

 

 

The same was true when we tried a few common SSRF protection bypass techniques like:

• Using a decimal value for the IP address (169.254.169.254 == 2852039166)
• Requesting a domain which resolves to 169.254.169.254 instead of requesting the IP address directly.

The next bypass we tried, which we had recently used against targets in Azure AI Studio and Azure ML Studio, and again in Azure Health Bot, was pointing the HttpRequestAction at a server we control, and sending a 301 redirect response that points to the restricted hosts.

Instead of receiving a System Error response like when requesting directly, using a controller we control to redirect requests to 169.254.169.254 resulted in a 400 error, which is progress. To make requests to IMDS, we need to be able to set the metadata header to “true”, which we were able to do using the tools in Copilot Studio. However there was the issue of an X-Forwarded-For header being added automatically. According to Microsoft documentation:

In order to ensure that requests are directly intended for IMDS and prevent unintended or unwanted redirection of requests, requests:

• Must contain the header Metadata: true
• Must not contain an X-Forwarded-For header

We needed to get rid of the X-Forwarded-For header if we wanted to be able to access the metadata service. Oddly enough, the Topic code editor allowed us to create multiline header values, which would allow us to insert newlines into any given header. 

 

 

If we added a metadata header, as well as some extra newlines at the end of the “true” value, we could push the X-Forwarded-For header to be part of the HTTP request body which would be ignored by the IMDS.

Combining this with the 301 redirect to 169.254.169.254, we were able to retrieve the instance metadata in a Copilot chat message:

 

While this information isn’t too sensitive on its own, we were also able to retrieve managed identity access tokens from the IMDS. We could then leverage this authentication token to access other internal resources. 

For example, redirecting Copilot to request http://169.254.169.254/metadata/identity/oauth2/token?api-version=2018-02-01&resource=https://management.azure.com/ would grant us an access token which lets us check whether this instance’s managed identity has access to any other Azure resources.

Discovering a Cosmos DB instance

It wasn’t immediately clear whether the access we’d already found was particularly impactful, whether it was limited to our environment, or if it impacted shared infrastructure and other tenants. With an access token for https://management.azure.com/ we can check to see if we have gained access to other Azure resources.

A good first check is to see which Azure subscriptions are associated with the identity by making a request to https://management.azure.com/subscriptions.  This revealed that the identity was associated with two subscriptions. Next we requested https://management.azure.com/subscriptions/<subscription-id>/resources to check for Azure resources we could potentially access.

On one of the two subscriptions we see a response like the following:

 

 

This tells us that there is a Cosmos DB resource available to us. Requesting details on the resource shows us the URLs for the Cosmos DB endpoints, and showed that we could access the Cosmos DB master keys which would grant us read/write permissions. 

Fortunately, this Cosmos DB instance is only accessible to IP addresses which belong to Microsoft infrastructure. Unfortunately, that range happened to include our Copilot.

Using Copilot to access the internal Cosmos DB

A quick review of some puzzle pieces we had and needed to put together:

• A URL for an internal Cosmos DB instance, accessible only to internal Azure infrastructure
• The master keys for that Cosmos DB instance
• The ability to make HTTP requests from Copilot, with control over request headers.

To access Cosmos DB we needed to provide Copilot with the document endpoint URL, which we obtained from management.azure.com, and appropriate authorization and x-ms-data headers, according to Microsoft’s documentation. 

Using the sample node.js code in the documentation above, along with the Cosmos DB keys, it was possible to generate a valid authorization token, along with the associated x-ms-date header. After providing all of the pieces to Copilot and sending the request we saw a valid response, indicating that we could leverage the SSRF vulnerability in Copilot to gain read/write access on this internal Cosmos DB instance.

 

 

Wrapping up

We tested this from multiple tenants and confirmed that, while no cross-tenant information appeared immediately accessible, the infrastructure used for this Copilot Studio service was shared among tenants. Any impact on that infrastructure could affect multiple customers. While we don’t know the extent of the impact that having read/write access to this infrastructure could have, it’s clear that because it’s shared among tenants, the risk is magnified.

We also determined we could access other internal hosts, unrestricted, on the local subnet to which our instance belonged (10.0.x.0/24).

Once reported, Microsoft responded quickly and began addressing the issue. Microsoft assigned this issue CVE-2024-38206 and communicated their evaluation to Tenable as a Critical Information Disclosure issue.

NIST has released the first encryption algorithms that can protect data against quantum attacks. Plus, MIT launched a new database of AI risks. Meanwhile, the CSA published a paper outlining the unique risks involved in building systems that use LLMs. And get the latest on Q2’s most prevalent malware, the Radar/Dispossessor ransomware gang and CVE severity assessments!

Dive into six things that are top of mind for the week ending August 16.

1 - NIST releases first quantum-resistant algos

Three encryption algorithms that can protect data from quantum computer attacks are ready, a significant step in the efforts to prevent a global data-theft disaster when these powerful systems become available around 2030. 

The post-quantum encryption standards are designed to secure “a wide range of electronic information,” including email messages and e-commerce transactions, the U.S. National Institute of Standards and Technology (NIST) announced this week.

Here’s the problem: Cybercriminals are eagerly waiting to get their hands on quantum computers, which will be able to decrypt data protected with existing public-key cryptographic algorithms.

 

 

While NIST is evaluating more post-quantum algorithms, the agency is urging system administrators to start transitioning to this first set of encryption tools right away because the integration process will take time.

“There is no need to wait for future standards,” said Dustin Moody, NIST mathematician and leader of the standardization project. “Go ahead and start using these three.”

These are the three encryption standards, called Federal Information Processing Standard (FIPS):

• FIPS 203 is intended to be the primary standard for general encryption. Its advantages include its speed of operation and its small encryption keys. 
• FIPS 204 is envisioned as the primary standard for protecting digital signatures. 
• FIPS 205 is also designed for digital signatures. 

The standards contain the encryption algorithms’ computer code, implementation instructions and their intended uses.

To get more details, check out:

• The announcement “NIST Releases First 3 Finalized Post-Quantum Encryption Standards”

• This video

   

For more information about the quantum computing cyberthreat:

• “CISA: Critical infrastructure must prep for quantum computing threat” (Tenable)
• “Think you can ignore quantum computing? Think again” (CIO)
• “Cybersecurity of Quantum Computing: A New Frontier” (Carnegie Mellon University)
• “Will Quantum Computers Become the Next Cyber-Attack Platform?” (InformationWeek)
• “Quantum computing could threaten cybersecurity measures. Here’s why – and how tech firms are responding” (World Economic Forum)

2 - MIT launches database of AI risks with 700-plus entries

The Massachusetts Institute of Technology (MIT) this week unveiled the AI Risk Repository, a database that aims to consolidate in a single place all risks associated with the use of artificial intelligence.

To compile the initial set of 700-plus risks, MIT analyzed 43 existing AI risk frameworks, and found that even the most comprehensive framework overlooked about 30% of all risks currently listed in the database.

“Since the AI risk literature is scattered across peer-reviewed journals, preprints, and industry reports, and quite varied, I worry that decision-makers may unwittingly consult incomplete overviews, miss important concerns, and develop collective blind spots,” project leader and MIT postdoctoral researcher Peter Slattery said in a statement.

 

 

The AI Risk Repository’s risk domains include:

• AI system safety, failures, and limitations
• Socioeconomic and environmental harms
• Discrimination and toxicity
• Privacy and security
• Malicious actors and misuse

The risk domains are further subdivided into 23 subdomains. The AI Risk Repository is a “living database” that’ll be expanded and updated, according to MIT.

To get more details, check out:

• The AI Risk Repository’s announcement
• The AI Risk Repository’s home page and FAQ

• This explainer video

   

For more information about AI risks:

• “4 Types of Gen AI Risk and How to Mitigate Them” (Harvard Business Review)
• “World leaders still need to wake up to AI risks” (Science Daily)
• “To understand the risks posed by AI, follow the money” (The Conversation)
• “Why can’t anyone agree on how dangerous AI will be?” (Vox)
• “Too Much Trust in AI Poses Unexpected Threats to the Scientific Process” (Scientific American)

3 - CSA: Beware of unique risks when building systems with LLMs

If your organization is building systems that use large language models (LLMs), your system architects and engineers need to be aware of special risks and challenges that impact security.

That’s the main topic of the Cloud Security Alliance’s new report “Securing LLM Backed Systems: Essential Authorization Practices,” published this week. It offers guidance and best practices for securely building systems that leverage LLMs.

“LLMs used as system components pose several new challenges for architects and engineers regarding predictability, security, and authorization,” reads the report, which is aimed at systems engineers and architects, and at privacy and security professionals.

The report also outlines essential components of systems that use LLMs:

• Vector databases, which are used for AI-system retrieval and processing
• Orchestrators, which coordinate LLM inputs and outputs
• LLM caches, which cache LLM content for faster retrieval, but need access controls
• Validators, which protect the input / output process against attacks like SQL injection, command injection and cross-site scripting

(Source: Cloud Security Alliance’s “Securing LLM Backed Systems: Essential Authorization Practices” report, August 2024)

To get more details, check out:

• The report’s announcement “Cloud Security Alliance Issues Paper Detailing Unique Challenges and Risks Presented by Use of Large Language Models (LLMs)”
• The full report “Securing LLM Backed Systems: Essential Authorization Practices”

For more information about AI risks to cybersecurity, check out these Tenable blogs:

• “How to Discover, Analyze and Respond to Threats Faster with Generative AI”
• “Never Trust User Inputs -- And AI Isn't an Exception: A Security-First Approach”
• “AI Is About To Take Cybersecurity By Storm: Here's What You Can Expect”
• “Securing the AI Attack Surface: Separating the Unknown from the Well Understood”
• “Do You Think You Have No AI Exposures? Think Again”

4 - A Tenable poll on CVE severity assessments

During a recent webinar about Tenable Security Center, we polled attendees about how they assess the severity of Common Vulnerabilities and Exposures (CVEs.) 

Check out what they said about their current and future use of the Common Vulnerability Scoring System (CVSS), the Vulnerability Priority Rating (VPR) and the Exploit Prediction Scoring System (EPSS).

(59 respondents polled by Tenable, August 2024. Respondents could choose more than one answer.)

(55 respondents polled by Tenable, August 2024. Respondents could choose more than one answer. CVSSv2 was an option but no one chose it.)

If you’re interested in the latest about Tenable Security Center and in vulnerability management best practices, watch the on-demand webinar “Tenable Security Center Customer Update, August 2024.”

5 - SocGholish again tops quarterly list of malware incidents

For the fourth straight time, the SocGholish malware variant – used in fake software-update attacks – ranked first on the Center for Internet Security’s quarterly report of malware infections.

SocGholish accounted for 60% of malware incidents in the second quarter of 2024, a sign that the popularity of fake software-update attacks remains strong. It first topped the list in the third quarter of 2023, with a 31% share of malware incidents.

In fake software-update attacks, users get tricked into installing a legitimate-looking update for, say, their preferred browser, that instead infects their computers with malware.

(Source: Center for Internet Security, August 2024)

Here’s the full list, in descending order:

• SocGholish, a downloader distributed through malicious websites that tricks users into downloading it by offering fake software updates 
• Agent Tesla, a remote access trojan (RAT) that captures credentials, keystrokes and screenshots
• CoinMiner, a cryptocurrency miner that spreads using Windows Management Instrumentation (WMI)
• NanoCore, a RAT that spreads via malspam as a malicious Excel spreadsheet
• ZPHP, a JavaScript downloader that’s distributed via fake browser updates
• Mirai, a malware botnet that compromises IoT devices to launch DDoS attacks
• Magecart, a credit card skimming and point-of-sale malware that steals payment data from forms on vulnerable websites 
• Arechclient2, also known as SectopRAT, is a .NET RAT whose capabilities include multiple stealth functions
• DarkGate, a downloader and keylogger that steals financial data and personally identifiable information (PII), and delivers other malware
• Lumma Stealer, an infostealer used to swipe PII, credentials, cookies and banking information

To get more information, check out the CIS blog “Top 10 Malware Q2 2024,” where you’ll find more details, context and indicators of compromise for each malware strain.

For details on fake update attacks:

• “The Fake Browser Update Scam Gets a Makeover” (Krebs on Security)
• “Watch Out: Attackers Are Hiding Malware in 'Browser Updates'” (Dark Reading)
• “Hackers Use Fake Browser Updates for AMOS Malware Attacks Targeting Mac Users” (MSSP Alert)
• “Malware crooks find an in with fake browser updates, in case real ones weren't bad enough” (The Register)
• “Hackers have been spreading malware via fake Chrome updates” (TechRadar)

VIDEOS

Fake Chrome Update Malware (The PC Security Channel)

Apple Mac OS users falling for fake browser update that installs malware (Windows, Computers and Technology)

6 - Radar/Dispossessor ransomware gang hit by int’l police operation

Law-enforcement agencies from Germany, the U.K. and the U.S. disrupted the Radar/Dispossessor ransomware group, shutting down its servers and domains in these three countries.

Radar/Dispossessor targets small- and medium-sized businesses (SMBs), looking for systems with unpatched vulnerabilities, weak passwords and no multi-factor authentication, the FBI announced this week. It uses a double-extortion model, exfiltrating victims’ data as well as encrypting it.

Although it initially targeted the U.S., law enforcement agencies found evidence that it has struck victims in about 15 other countries in Europe, Asia and Latin America. Launched in August 2023, it is led by an individual known as “Brain,” according to the FBI.

Ivanti released a patch for a critical severity authentication bypass vulnerability and a warning that exploit code is publicly available

Background

On August 13, Ivanti released a security advisory to address a critical severity authentication bypass vulnerability in its Virtual Traffic Manager (vTM) product, a software-based application delivery controller (ADC).

CVEDescriptionCVSSv3CVE-2024-7593Ivanti Virtual Traffic Manager (vTM) Authentication Bypass Vulnerability9.8Analysis

CVE-2024-7593 is a critical severity authentication bypass vulnerability. Exploitation of this flaw could allow an unauthenticated, remote attacker to create an administrator user. According to the advisory, access to the management interface of vTM is required in order to exploit this vulnerability.

While Ivanti notes that no known exploitation has been observed for CVE-2024-7593, their advisory makes special note of publicly available exploit code.

Historical exploitation of Ivanti vulnerabilities

Ivanti has had a history of threat actors targeting their products, with one of the most prolific being attacks against their Ivanti Connect Secure (ICS) product, previously known as Pulse Connect Secure and Ivanti Policy Secure. In January, two zero-day vulnerabilities (CVE-2023-46805, CVE-2024-21887) were exploited against ICS in a chained attack. Coincidentally enough, one of these vulnerabilities (CVE-2023-46805) was an authentication bypass flaw. Within a few weeks of this release, Ivanti found several additional vulnerabilities impacting ICS, including another zero-day vulnerability (CVE-2024-21893) that had also been exploited in the wild. CVE-2023-46805 and CVE-2024-21887 were reportedly exploited by a threat actor tracked as UTA0178 and is believed to be a “Chinese nation-state level threat actor.” With threat actors, including nation-state aligned actors actively targeting Ivanti devices, it’s imperative that patching is prioritized.

Hundreds of Internet Facing Instances May Be Affected

Adding to the concern over the potential to exploit this vulnerability, a FOFA search lists more than 400 results tied to over 200 unique IPs that might be affected if patches or mitigations are not applied.

Image source: https://en.fofa.info/

Proof of concept

According to Ivanti’s security advisory, a public proof-of-concept (PoC) is available for this vulnerability. While Tenable Research has not tested and confirmed a working exploit, we are aware of exploit code found on a popular site for public exploit scripts. As attackers are keen to abuse targets with readily available exploit code, immediate patching or application of mitigation steps is recommended.

Solution

Ivanti has released some patches to address this vulnerability, with additional patch versions expected to be released the week of August 19. A summary of the patched versions can be found in the table below:

Affected VersionFixed VersionAvailability22.222.2R1Available now22.322.3R3Upcoming- Week of August 1922.3R222.3R3Upcoming- Week of August 1922.5R122.5R2Upcoming- Week of August 1922.6R122.6R2Upcoming- Week of August 1922.7R122.7R2Available now

For organizations that are not able to immediately patch this vulnerability, Ivanti does offer mitigation guidance. As access to the management interface is a requirement for exploitation, limiting access to this interface is recommended. Ivanti recommends to ensure that the management interface is bound to an internal network or private IP address. We recommend referring to Ivanti’s security advisory for the most up to date information on patch availability and mitigation guidance.

Identifying affected systems

A list of Tenable plugins for this vulnerability can be found on the individual CVE page for CVE-2024-7593 as they’re released. This link will display all available plugins for this vulnerability, including upcoming plugins in our Plugins Pipeline.

Get more information

• Ivanti Security Advisory: Ivanti Virtual Traffic Manager (vTM ) (CVE-2024-7593)

Join Tenable's Security Response Team on the Tenable Community.

Learn more about Tenable One, the Exposure Management Platform for the modern attack surface.

1. 7Critical
2. 80Important
3. 1Moderate
4. 0Low

Microsoft addresses 88 CVEs with seven critical vulnerabilities and 10 zero-day vulnerabilities, six of which were exploited in the wild.

Microsoft patched 88 CVEs in its August 2024 Patch Tuesday release, with seven rated critical, 80 rated as important, and one rated as moderate.

This month’s update includes patches for:

• .NET and Visual Studio
• Azure Connected Machine Agent
• Azure CycleCloud
• Azure Health Bot
• Azure IoT SDK
• Azure Stack
• Line Printer Daemon Service (LPD)
• Microsoft Bluetooth Driver
• Microsoft Copilot Studio
• Microsoft Dynamics
• Microsoft Edge (Chromium-based)
• Microsoft Local Security Authority Server (lsasrv)
• Microsoft Office
• Microsoft Office Excel
• Microsoft Office Outlook
• Microsoft Office PowerPoint
• Microsoft Office Project
• Microsoft Office Visio
• Microsoft Streaming Service
• Microsoft Teams
• Microsoft WDAC OLE DB provider for SQL
• Microsoft Windows DNS
• Reliable Multicast Transport Driver (RMCAST)
• Windows Ancillary Function Driver for WinSock
• Windows App Installer
• Windows Clipboard Virtual Channel Extension
• Windows Cloud Files Mini Filter Driver
• Windows Common Log File System Driver
• Windows Compressed Folder
• Windows Deployment Services
• Windows DWM Core Library
• Windows Initial Machine Configuration
• Windows IP Routing Management Snapin
• Windows Kerberos
• Windows Kernel
• Windows Kernel-Mode Drivers
• Windows Layer-2 Bridge Network Driver
• Windows Mark of the Web (MOTW)
• Windows Mobile Broadband
• Windows Network Address Translation (NAT)
• Windows Network Virtualization
• Windows NT OS Kernel
• Windows NTFS
• Windows Power Dependency Coordinator
• Windows Print Spooler Components
• Windows Resource Manager
• Windows Routing and Remote Access Service (RRAS)
• Windows Scripting
• Windows Secure Kernel Mode
• Windows Security Center
• Windows SmartScreen
• Windows TCP/IP
• Windows Transport Security Layer (TLS)
• Windows Update Stack
• Windows WLAN Auto Config Service

Elevation of Privilege (EoP) vulnerabilities accounted for 41% of the vulnerabilities patched this month, followed by Remote Code Execution (RCE) at 33%.

CriticalCVE-2024-38206 | Microsoft Copilot Studio Information Disclosure Vulnerability

CVE-2024-38206 is a critical severity information disclosure vulnerability affecting Microsoft’s Copilot Studio, an AI-powered chatbot. This vulnerability received a CVSSv3 score of 8.5 and could be abused by an authenticated attacker to bypass server-side request forgery (SSRF) protections in order to leak potentially sensitive information. The vulnerability was released by Microsoft on August 6, with the advisory noting that no user action is required as the issue has been patched by Microsoft. This vulnerability was discovered and reported to Microsoft by Tenable researcher Evan Grant.

CriticalCVE-2024-38109 | Azure Health Bot Elevation of Privilege Vulnerability

CVE-2024-38109 is a critical severity EoP vulnerability affecting Azure Health Bot. This vulnerability received a CVSSv3 score of 9.1 and is the result of a SSRF vulnerability in Azure Health Bot that can be abused to escalate privileges. This vulnerability was discovered by Tenable researcher Jimi Sebree and responsibly disclosed to Microsoft. The issue has been patched by Microsoft and no action is required for users of the Health Bot service. For more information on this vulnerability, please refer to Tenable Research Advisories TRA-2024-27 and TRA-2024-28, as well as our blog post.

ImportantCVE-2024-38106, CVE-2024-38133 and CVE-2024-38153 | Windows Kernel Elevation of Privilege Vulnerability

CVE-2024-38106, CVE-2024-38133 and CVE-2024-38153 are EoP vulnerabilities affecting the Windows Kernel. CVE-2024-38133 and CVE-2024-38153 received CVSSv3 scores of 7.8, while CVE-2024-38106 was scored as a 7. Despite the lower severity and the exploitability requirements of the attacker needing to win a race condition for successful exploitation, CVE-2024-38106 was reportedly exploited in the wild as a zero-day. CVE-2024-38133 and CVE-2024-38153 were not listed as being exploited, however CVE-2024-38133 was rated as “Exploitation More Likely” according to the Microsoft Exploitability Index. Successful exploitation of these vulnerabilities could allow the attacker to elevate privileges to SYSTEM.

ImportantCVE-2024-38107 | Windows Power Dependency Coordinator Elevation of Privilege Vulnerability

CVE-2024-38107 is an EoP Vulnerability affecting Windows Power Dependency Coordinator (pdc.sys), a driver responsible for power management on a Windows system. This vulnerability was exploited in the wild as a zero-day, though no specific details about exploitation were available at the time this blog was published. Microsoft provided a CVSSv3 score of 7.8 for this vulnerability and patches are available for all supported versions of Windows and Windows Server.

ImportantCVE-2024-38178 | Scripting Engine Memory Corruption Vulnerability

CVE-2024-38178 is a Scripting Engine memory corruption vulnerability in Windows Scripting. This vulnerability was assigned a CVSSv3 score of 7.5 and Microsoft notes that exploitation has been observed. According to Microsoft, an authenticated victim must have Edge in Internet Explorer Mode as a prerequisite for exploitation prior to an unauthenticated attacker convincing the victim to click a specially crafted URL to obtain RCE.

ImportantCVE-2024-38189 | Microsoft Project Remote Code Execution Vulnerability

CVE-2024-38189 is a RCE vulnerability affecting Microsoft Project, a project management tool. This vulnerability received a CVSSv3 score of 8.8 and was exploited in the wild. According to the advisory, exploitation requires an unsuspecting victim to open a crafted Microsoft Office Project file. Additionally, the system must be configured to have the “Block macros from running in Office files from the Internet policy” disabled as well as have the VBA Macro Notification Settings disabled in order for a successful attack. Micsoft’s advisory does clarify that the Preview Pane is not an attack vector for this vulnerability and offers mitigation options to protect systems if immediate patching cannot be immediately performed.

ImportantCVE-2024-38141 and CVE-2024-38193 | Windows Ancillary Function Driver for WinSock Elevation of Privilege Vulnerability

CVE-2024-38141 and CVE-2024-38193 are EoP vulnerabilities affecting the Windows Ancillary Function Driver for Winsock (afd.sys). Both of these vulnerabilities were given CVSSv3 scores of 7.8 and can allow an attacker to escalate privileges to SYSTEM. CVE-2024-38141 is rated as “Exploitation More Likely” and CVE-2024-38193 was reported to have been exploited in the wild as a zero-day vulnerability.

ModerateCVE-2024-38213 | Windows Mark of the Web Security Feature Bypass Vulnerability

CVE-2024-38213 is a security feature bypass vulnerability with an assigned CVSSv3 score of 6.5. Exploitation of this vulnerability requires a user to open a specially crafted file, which could be hosted on a file server, website or sent via a phishing email. If the attacker is successful in convincing a victim to open this file, they could bypass the Windows SmartScreen user experience. Microsoft has flagged this as “Exploitation Detected” as they are aware of an instance of this vulnerability being exploited.

ImportantCVE-2024-38163 and CVE-2024-38202 | Windows Update Stack Elevation of Privilege Vulnerability

CVE-2024-38163 and CVE-2024-38202 are both EoP vulnerabilities in Windows Update Stack and were assigned CVSSv3 scores of 7.8 and 7.3 respectively. CVE-2024-38163, if successfully exploited could result in gaining SYSTEM privileges. Microsoft has noted that users don’t need to take any action for this vulnerability as it is only exploitable at run time and the impacted version of WinRE has been superseded by a new version.

CVE-2024-38202 was disclosed and presented at BlackHat USA 2024 and DEF CON 32 by SafeBreach Labs researcher Alon Leviev prior to the August 2024 Patch Tuesday release. The vulnerability which exists in Windows Backup, allows a user with basic privileges to “reintroduce previously mitigated vulnerabilities or circumvent some features of Virtualization Based Security (VBS)”. Leviev identified the vulnerability in the Windows Update mechanism that could allow unauthorized elevation of privileges by enforcing the downgrade of system components. This vulnerability exposes systems to previously patched exploits, making them susceptible to attacks that could leverage these old vulnerabilities. Microsoft has noted that “an attacker attempting to exploit this vulnerability requires additional interaction by a privileged user to be successful.” Microsoft issued an advisory in coordination with this disclosure at Black Hat.

Source: Leviev's SafeBreach Blog

ImportantCVE-2024-21302 and CVE-2024-38142 | Windows Secure Kernel Mode Elevation of Privilege Vulnerability

CVE-2024-21302 and CVE-2024-38142 are both elevation of privilege vulnerabilities in Windows Secure Kernel with an exploitability assessment by Microsoft as “Exploitation Less Likely”. CVE-2024-21302 carries a CVSSv3 score of 6.7 and CVE-2024-38142 a score of 7.8 with successful exploitation of either of these vulnerabilities resulting in an attacker gaining SYSTEM privileges.

CVE-2024-21302 was disclosed at Black Hat USA 2024 by the previously mentioned security researcher, Alon Leviev. Leviev demonstrated that CVE-2024-21302 could be chained with CVE-2024-38202 to downgrade or roll back software versions without the need for interaction from a victim with elevated privileges. The result of this chained attack is the target device could be made susceptible to previously patched vulnerabilities, increasing the attack surface of the device. CVE-2024-21302 was also included in the previously mentioned Microsoft advisory released in coordination with the disclosure at Black Hat.

ImportantCVE-2024-38199 | Windows Line Printer Daemon (LPD) Service Remote Code Execution Vulnerability

CVE-2024-38199 is a RCE vulnerability in Windows Line Printer Daemon (LPD) Service. The flaw was assigned a CVSSv3 score of 9.8 and rated “Exploitation Less Likely” by Microsoft. A remote attacker could exploit this across a network by dispatching a specially crafted print task to Windows LPD Service, if successful it would result in RCE on the server. Microsoft has also noted that it was publicly disclosed prior to a patch being available.

ImportantCVE-2024-38200 | Microsoft Office Spoofing Vulnerability

CVE-2024-38200 is a spoofing vulnerability affecting Microsoft Office with a CVSSv3 score of 6.5 and rated by Microsoft as “Exploitation Less Likely”. An attacker could leverage this vulnerability with a specially crafted file that a victim would need to interact with. This could be achieved by hosting it on a file server or website and convincing the victim to click on the file or similarly it could be included in a phishing email. Successful exploitation of the vulnerability could result in the victim exposing NTLM (New Technology Lan Manager) hashes to a remote attacker.

CVE-2024-38200 was publicly disclosed on August 8 at DEF CON 32 by Jim Rush and Tomais Williamson, both Senior Security Consultants at PrivSec Consulting. Patches for this vulnerability were released today as part of the August 2024 Patch Tuesday released, however on August 10, Microsoft provided mitigations in coordination with this disclosure prior to the patch being released.

CriticalCVE-2024-38063 | Windows TCP/IP Remote Code Execution Vulnerability

CVE-2024-38063 is a critical RCE vulnerability affecting Windows TCP/IP. It received a CVSSv3 score of 9.8 and is rated as “Exploitation More Likely.” An attacker could remotely exploit this vulnerability by sending specially crafted IPv6 packets to a host. Microsoft’s mitigation suggestions suggest disabling IPv6 as only IPv6 packets can be abused to exploit this vulnerability. Microsoft has released patches for all supported versions of Windows and Windows Server, including Server Core installations.

Tenable Solutions

A list of all the plugins released for Microsoft’s August 2024 Patch Tuesday update can be found here. As always, we recommend patching systems as soon as possible and regularly scanning your environment to identify those systems yet to be patched.

For more specific guidance on best practices for vulnerability assessments, please refer to our blog post on How to Perform Efficient Vulnerability Assessments with Tenable.

Get more information

• Microsoft's August 2024 Security Updates
• Tenable plugins for Microsoft August 2024 Patch Tuesday Security Updates

Join Tenable's Security Response Team on the Tenable Community.

Learn more about https://www.tenable.com/products/tenable-one">Tenable One, the Exposure Management Platform for the modern attack surface.

Tenable Research discovered multiple privilege-escalation issues in the Azure Health Bot Service via a server-side request forgery (SSRF), which allowed researchers access to cross-tenant resources.

Key takeaways

• The Azure Health Bot Service is a cloud platform that allows healthcare professionals to deploy AI-powered virtual health assistants.
• Tenable Research discovered critical vulnerabilities that allowed access to cross-tenant resources within this service. Based on the level of access granted, it’s likely that lateral movement to other resources would have been possible.
• According to Microsoft, mitigations for these issues have been applied to all affected services and regions. No customer action is required. 

(Source: Image generated via ChatGPT 4o / DALL-E by Nick Miles)

An overview of the Azure Health Bot Service

This is how Microsoft describes the Azure Health Bot Service:

“The Azure Health Bot Service is a cloud platform that empowers developers in Healthcare organizations to build and deploy compliant, AI-powered virtual health assistants, that help them improve processes and reduce costs. It allows healthcare organizations to create experiences that act as copilots for their healthcare professionals to further manage administrative workloads, and experiences that engage with their patients.”

Essentially, the service allows healthcare providers to create and deploy patient-facing chatbots to handle administrative workflows within their environments. Thus, these chatbots generally have some amount of access to sensitive patient information, though the information available to these bots can vary based on each bot’s configuration.

While auditing this service for security issues, Tenable researchers became interested in a feature dubbed "Data Connections" in the service's documentation. These data connections allow bots to interact with external data sources to retrieve information from other services that the provider may be using, such as a portal for patient information or a reference database for general medical information.

The first discovery

This data connection feature is designed to allow the service’s backend to make requests to third-party APIs. While testing these data connections to see if endpoints internal to the service could be interacted with, Tenable researchers discovered that many common endpoints, such as Azure’s Internal Metadata Service (IMDS), were appropriately filtered or inaccessible. Upon closer inspection, however, it was discovered that issuing redirect responses (e.g. 301/302 status codes) allowed these mitigations to be bypassed.

“A server-side request forgery is a web security vulnerability that allows an attacker to force an application on a remote host to make requests to an unintended location.”

For example, by configuring a data connection within the service’s scenario editor, researchers were able to specify an external host under their control.

On this external host, researchers configured it to respond to requests with a 301 redirect response destined for Azure’s IMDS.

#!/usr/bin/python3 from http.server import HTTPServer, BaseHTTPRequestHandler def servePage(s, hverb): s.protocol_version = 'HTTP/1.1' s.server_version = 'Microsoft-IIS/8.5' s.sys_version = '' s.send_response(301) s.send_header('Location','http://169.254.169.254/metadata/instance?api-version=2021-12-13') s.end_headers() message = "" s.wfile.write(bytes(message, "utf8")) return class StaticServer(BaseHTTPRequestHandler): def do_GET(self): servePage(self, "GET") return def main(server_class=HTTPServer, handler_class=StaticServer, port=80): server_address = ('', port) httpd = server_class(server_address, handler_class) httpd.serve_forever() main()

After obtaining a valid metadata response, researchers attempted to obtain an access token for management.azure.com.

With this token, researchers were then able to list the subscriptions they had access to via a call to https://management.azure.com/subscriptions?api-version=2020-01-01, which provided them with a subscription ID internal to Microsoft.

Finally, researchers were able to list the resources they had access to via https://management.azure.com/subscriptions/<REDACTED>/resources?api-version=2020-10-01. The resulting list of resources contained hundreds and hundreds of resources belonging to other customers.

A quick fix

Upon seeing that these resources contained identifiers indicating cross-tenant information (i.e. information for other users/customers of the service), Tenable researchers immediately halted their investigation of this attack vector and reported their findings to MSRC on June 17, 2024. MSRC acknowledged Tenable’s report and began their investigation the same day. 

Within the week, MSRC confirmed Tenable’s report and began introducing fixes into the affected environments. As of July 2, MSRC has stated that fixes have been rolled out to all regions. To Tenable’s knowledge, no evidence was discovered that indicated this issue had been exploited by a malicious actor.

Doing it again

Once MSRC stated that this issue had been fixed, Tenable Research picked up where it left off to confirm that the original proof-of-concepts provided to MSRC during the disclosure process were no longer functional. As it turns out, the fix for this issue was to simply reject redirect status codes altogether for data connection endpoints, which eliminated this attack vector.

That said, researchers discovered another endpoint used for validating data connections for FHIR endpoints. This validation mechanism was more or less vulnerable to the same attack described above. The difference between this issue and the first is the overall impact. The FHIR endpoint vector did not have the ability to influence request headers, which limits the ability to access IMDS directly. While other service internals are accessible via this vector, Microsoft has stated that this particular vulnerability had no cross-tenant access.

As before, the researchers immediately halted their investigation and reported the finding to Microsoft, opting to respect MSRC’s guidance regarding accessing cross-tenant resources. This second issue was reported on July 9 with fixes available by July 12. As with the first issue, to Tenable’s knowledge, no evidence was discovered that indicated this issue had been exploited by a malicious actor.

Conclusion

The vulnerabilities discussed in this post involve flaws in the underlying architecture of the AI chatbot service rather than the AI models themselves. As explained by Lucas Tamagna-Darr in a Tenable research blog post last week, this highlights the continued importance of traditional web application and cloud security mechanisms in this new age of AI powered services.

Please see TRA-2024-27 and TRA-2024-28 for more information regarding each of the discoveries mentioned in this post.

The Cloud Security Alliance has released its list of top cloud threats for 2024. Plus, CISA and the FBI published a guide for determining if a software product was built "secure by design." Meanwhile, find out how AI can transform offensive security. And get the latest on the Royal ransomware gang, the CIS Benchmarks and TikTok’s legal troubles!

Dive into six things that are top of mind for the week ending August 9.

1 - Report: Top cloud security threats for 2024

Misconfigurations. Identity weaknesses. Insecure APIs. Incomplete security strategies.

Those are the top four dangers impacting cloud environments today, according to the Cloud Security Alliance’s “Top Threats to Cloud Computing 2024” report, based on a survey of 500-plus cloud security experts.

“By bringing attention to those threats, vulnerabilities, and risks that are top-of-mind across the industry, organizations can better focus their resources,” Sean Heide, Technical Research Director at the CSA, said in a statement.

The report lists a total of 11 major cloud-computing threats, describes and analyzes them, identifies their business impacts, offers key takeaways, provides real-world examples and more. 

Here’s the full list:

• Misconfiguration and inadequate change control
• Identity and access management
• Insecure interfaces and APIs
• Inadequate selection/implementation of cloud security strategy
• Insecure third-party resources
• Insecure software development
• Accidental cloud disclosure
• System vulnerabilities
• Limited cloud visibility/observability
• Unauthenticated resource sharing
• Advanced persistent threats

The report also outlines four critical trends that will make cloud security more challenging in the future:

• Increasingly sophisticated attacks, thanks to cybercriminals’ use of technologies like AI
• Deeper supply chain risk, caused by the growing complexity of cloud ecosystems
• Stricter regulations for data privacy and security
• A democratization of cyberattack capabilities due to the prevalence of ransomware-as-a-service alternatives

So, what can security teams do? Recommendations include:

• Leverage AI throughout the software development lifecycle (SDLC) to improve code reviews and vulnerability scanning, and thus reduce security issues in released software.
• Use AI tools that simulate cyberattacks to detect security gaps in cloud configurations, IAM systems and APIs.
• Use cloud-native security tools that have been designed specifically for cloud environments.
• Adopt a Zero Trust security model.
• Incorporate security automation and orchestration technologies to tackle cloud complexity.
• Close your organization’s security skills gap by investing in cloud security training and awareness.

To get more details, check out:

• The report’s announcement, “Cloud Security Alliance Releases Top‌ ‌Threats‌ ‌to‌ ‌Cloud‌ ‌Computing 2024 Report”
• The full report “Top Threats to Cloud Computing 2024”

For more information about cloud security, check out these Tenable resources:

• “Empower Your Cloud: Mastering CNAPP Security” (eBook)
• “Improving Your Cloud Security Using JIT Access for Sensitive SaaS Applications” (blog)
• “Tenable Cloud Security Outlook 2024” (on-demand webinar)
• “Understanding Customer Managed Encryption Keys (CMKs) in AWS, Azure and GCP: A Comparative Insight” (blog)
• “Secure Your Cloud-Native Applications: 5 Key Considerations” (on-demand webinar)

2 - New guide helps orgs assess the security of software products

During your organization’s software-procurement process, a critical evaluation criteria should be whether the products under consideration are secure by design. But how do you make that assessment?

Check out new guidance to help organizations make sure that they buy secure-by-design software, which is software whose security was prioritized by its manufacturer throughout the product’s development process.

“Ensuring that the products they use and procure are secure by design is essential for organizations to be resilient against ransomware and other forms of malicious cyber activity,” reads the guidance, titled “Secure by Demand Guide: How Software Customers Can Drive a Secure Technology Ecosystem.”

Published this week by the U.S. Cybersecurity and Infrastructure Security Agency (CISA) and the FBI, the guidance includes:

• Security-related questions to ask software manufacturers
• Ways to assess a software product’s security throughout the procurement process
• Resources to determine product-security maturity

Among the topics covered are:

• Does the product offer secure authentication, such as by supporting phishing-resistant forms of authentication?
• Has the manufacturer eliminated well-known classes of vulnerabilities and defects from its software by, for example, using memory-safe languages and enforcing parameterized queries?
• Does the manufacturer make security logs available to customers?
• Does the manufacturer keep and share provenance data of the third-party components it uses in its software?
• Does the manufacturer quickly and transparently disclose vulnerabilities found in its software?

Ultimately, a key goal is to empower organizations to leverage their purchasing power to procure secure software products, turning the “secure by design” principle into “secure by demand,” CISA Director Jen Easterly said in a statement.

To get more details:

• Check out the announcement “CISA Releases Secure by Demand Guide”
• Read the full guide “Secure by Demand Guide: How Software Customers Can Drive a Secure Technology Ecosystem”

3 - Study: AI has potential to transform offensive security

Although challenges remain, AI holds great promise for offensive-security teams, especially those involved with vulnerability assessments, penetration testing and red teaming.

That’s the conclusion from the Cloud Security Alliance’s “Using AI for Offensive Security” study, published this week.

Offensive-security teams, tasked with identifying their organizations’ cybersecurity weak spots, are already benefiting from AI in general, and from large language models (LLMs) and LLM-powered AI agents in particular. Specifically, these teams are seeing increased speed, automation, data analysis, scalability and productivity.

“This improvement boosts efficiency, allows for more sophisticated and extensive assessments, and enables security teams to focus on process improvement and strategic work,” the 29-page study reads.

An added bonus: AI helps understaffed cybersecurity teams because it lowers the barriers to entry to offensive-security teams and democratizes security testing, according to the CSA. In fact, the cybersecurity skills shortage is one of the challenges the study highlights.

The study explores how AI can help teams across the five phases of offensive-security testing:

• Reconnaissance: Gathering as much information as possible about the IT environment
• Scanning: Probing for details such as live hosts, open ports and running services
• Vulnerability analysis: Identifying security weaknesses in the IT environment
• Exploitation: Exploiting identified vulnerabilities to gain initial access or move deeper into the IT environment
• Reporting: Compiling all findings into a detailed report

How AI can augment or automate an existing offensive-security testing process

(Source: Cloud Security Alliance’s “Using AI for Offensive Security” report, August 2024)

Recommendations include:

• Incorporate AI for task automation and for augmenting human capabilities, and leverage it for data analysis, tool orchestration, generating actionable insights and more.
• Adopt robust governance, risk and compliance (GRC) frameworks to ensure you’re using AI securely and responsibly.

To get more details, check out:

• The study’s announcement “Cloud Security Alliance Addresses Using Artificial Intelligence (AI) for Offensive Security in New Report”
• The full study “Using AI for Offensive Security”

To learn more about some of the ways in which AI and cybersecurity intersect, check out these Tenable blogs:

• “How to Discover, Analyze and Respond to Threats Faster with Generative AI”
• “Never Trust User Inputs -- And AI Isn't an Exception: A Security-First Approach”
• “AI Is About To Take Cybersecurity By Storm: Here's What You Can Expect”
• “Securing the AI Attack Surface: Separating the Unknown from the Well Understood”
• “Do You Think You Have No AI Exposures? Think Again”

4 - Advisory: Royal ransomware gang now goes by BlackSuit

The Royal ransomware group has changed its name to BlackSuit and revamped its tactics, techniques and procedures (TTPs) to sharpen its attacks, CISA and the FBI announced this week in an updated advisory about this cybercrime posse.

The joint advisory, first published in March 2023, is now titled “#StopRansomware: BlackSuit (Royal) Ransomware” and also includes new indicators of compromise (IOCs), as well as new detection methods. The advisory’s most recent data was obtained as recently as July 2024.

Here’s a sampling of new and updated information:

• Before encrypting victims’ data, BlackSuit exfiltrates the data and engages in extortion tactics. If victims don’t pay the ransom, BlackSuit publishes their data to a leak site.
• Phishing emails are one of BlackSuit’s preferred methods for gaining initial access to a victim’s network. 
• BlackSuit’s ransom demands range from $1 million to $10 million. It requires that payment be made using Bitcoin. 
• To date, BlackSuit attackers have demanded $500 million-plus in ransoms, and have shown a willingness to negotiate payment amounts. Attackers prefer to communicate ransom demands via telephone calls and email messages.

Mitigation recommendations include:

• Make it a priority to patch known vulnerabilities that have been exploited.
• Teach users how to recognize phishing attacks.
• Protect admin accounts with phishing-resistant multi-factor authentication and with time-based access.
• Segment networks to prevent ransomware infections from spreading.
• Back up data offline and encrypt it.

To get more details, check out:

• CISA’s alert “Royal Ransomware Actors Rebrand as “BlackSuit,” FBI and CISA Release Update to Advisory”
• The updated joint advisory “#StopRansomware: BlackSuit (Royal) Ransomware”

5 - U.S. DoJ sues TikTok over children’s privacy

The U.S. government has sued TikTok and its parent company ByteDance for allegedly violating the Children’s Online Privacy Protection Act (COPPA).

In a civil lawsuit filed last week in district court, the U.S. government alleges that TikTok has knowingly allowed children under 13 to create regular accounts and use the social media service to create and share videos, and to exchange messages with adults.

The complaint, filed in U.S. District Court for the Central District of California by the Department of Justice and the Federal Trade Commission, further alleges that TikTok and ByteDance:

• Collected and stored personal information from these children without alerting nor getting consent from their parents
• Frequently failed to honor parents’ requests to delete their children’s accounts and information
• Had “deficient and ineffectual” policies and processes for identifying and deleting accounts created by children
• Unlawfully collected and retained children's personal information even in "Kids Mode" accounts, which are tailored for users under 13

TikTok and ByteDance have been under a court order barring them from violating COPPA since 2019, when TikTok’s predecessor Musical.ly settled a lawsuit with the U.S. government. This new lawsuit seeks civil penalties and injunctive relief.

“With this action, the Department seeks to ensure that TikTok honors its obligation to protect children’s privacy rights and parents’ efforts to protect their children,” Acting Associate Attorney General Benjamin C. Mizer said in a statement.

In comments to the media, TikTok disputed the U.S. government’s claims. “We disagree with these allegations, many of which relate to past events and practices that are factually inaccurate or have been addressed,” TikTok spokesperson Michael Hughes said in a statement sent to CNN.com.

President Biden signed a bill into law that requires ByteDance to sell its U.S. TikTok operations by January 19, 2025. If it doesn’t, TikTok would be banned in the U.S. ByteDance is challenging the law in a U.S. appeals court.

For more information about privacy and security concerns around TikTok:

• “Is TikTok a National Security Risk?” (The New York Times)
• “Majority of Americans say TikTok is a threat to national security” (Pew Research Center)
• “TikTok Ban Raises Data Security, Control Questions” (Dark Reading)
• “FTC investigating TikTok over privacy and security” (CNN)
• “How TikTok Grew From a Fun App for Teens Into a Potential National Security Threat” (Associated Press)

VIDEOS

DOJ suing TikTok (ABC News)

The data security concerns surrounding social media app TikTok (The Financial Times)

6 - CIS releases new or updated Benchmarks for Microsoft, Google and Mozilla products

The Center for Internet Security issued new and updated CIS Benchmarks for various products, including Microsoft Office Enterprise, Google Kubernetes Engine (GKE) Autopilot and Mozilla Firefox Extended Support Release (ESR) Group Policy (GPO).

Here’s the full list of updated and new CIS Benchmarks for July.

Updated

• CIS Apache HTTP Server 2.4 Benchmark v2.2.0
• CIS Docker Benchmark v1.7.0
• CIS Microsoft Edge Benchmark v3.0.0
• CIS Microsoft Office Enterprise Benchmark v1.2.0

New

• CIS Google Kubernetes Engine (GKE) Autopilot Benchmark v1.0.0
• CIS Mozilla Firefox ESR GPO Benchmark v1.0.0
• CIS Talos Linux Benchmark v1.0.0

Organizations use the CIS Benchmarks’ secure-configuration guidelines to harden products against attacks. Currently, CIS offers 100-plus Benchmarks for 25-plus vendor product families. Categories of products include cloud platforms; desktop and server software; mobile devices; operating systems; and more.

To get more details, read the CIS blog “CIS Benchmarks August 2024 Update.” For more information about the CIS Benchmarks list, check out its home page, as well as:

• “How to use CIS benchmarks to improve public cloud security” (TechTarget)
• “How to Unlock the Security Benefits of the CIS Benchmarks” (Tenable)
• “Getting to Know the CIS Benchmarks” (CIS)
• “CIS Benchmarks Communities: Where configurations meet consensus” (HelpNet Security)
• “CIS Benchmarks: DevOps Guide to Hardening the Cloud” (DevOps)

VIDEO

CIS Benchmarks (CIS)

 

###

Critical vulnerability in Cisco Smart Software Manager On-Prem exposes systems to unauthorized password changes, exploit code now available.

Background

On July 17, 2024, Cisco published an advisory for a critical vulnerability in Cisco’s Smart Software Manager On-Prem (SSM On-Prem):

CVEDescriptionCVSSv3CVE-2024-20419Cisco Smart Software Manager On-Prem Password Change Vulnerability10Analysis

CVE-2024-20419 is an unverified password change weakness within the Cisco SSM On-Prem interface due to improper validation. Specifically, the flaw allows an unauthenticated, remote attacker to exploit an insufficient authentication mechanism, changing the password of any user by sending specially crafted HTTP requests without prior knowledge of the existing password. The vulnerability affects Cisco SSM On-Prem version 8-202206 and earlier, including releases prior to 7.0 where the product was named Cisco SSM Satellite.

Successful exploitation could result in access to the web interface or API of Cisco SSM On-Prem in the context of the compromised user account. The vulnerability is considered critical as the complexity of the attack is low and could lead to full administrative control over the SSM On-Prem instance. This control could be used to disrupt the organization's software management processes, gain unauthorized access to sensitive resources and potentially carry out further attacks within the network.

On August 7, 2024, Cisco updated their advisory to reflect that public proof-of-concept (PoC) exploit code was now available, heightening the urgency to patch.

Proof of concept

On July 20, 2024, Mohammed Adel, a penetration tester with a previous history of developing exploits, published a detailed writeup explaining the root cause of the vulnerability and his PoC exploit code. The writeup also visually demonstrates the PoC being leveraged against an administrative account to change the password highlighting the vulnerabilities ease of exploitation.

Credit: Mohammeds writeup

Solution

Cisco has issued patches for all affected versions of Cisco SSM On-Prem. At the time of this blog there is no indication of exploitation in-the-wild, but with exploit code publicly available, administrators are strongly advised to apply these patches without delay to mitigate the risk of exploitation. The following table reflects affected and patched versions:

Affected VersionsPatched Version8-202206 and earlier8-2022129Not Vulnerable

Cisco has highlighted that there are no workarounds for this vulnerability.

Identifying affected systems

A list of Tenable plugins for this vulnerability can be found on the individual CVE page for CVE-2024-20419 as they’re released. This link will display all available plugins for this vulnerability, including upcoming plugins in our Plugins Pipeline.

Get more information

• Security Advisory: Cisco Smart Software Manager On-Prem Password Change Vulnerability
• Mohammed Adel Blog: Cisco SSM On-Prem; Account Takeover (CVE-2024-20419)

Join Tenable's Security Response Team on the Tenable Community.

Learn more about Tenable One, the Exposure Management Platform for the modern attack surface.

Kernel-mode drivers are critical yet risky components of the Windows operating system. Learn about their functionality, the dangers they pose, and how Tenable's new plugins can help identify and mitigate vulnerabilities using community-driven resources like LOLDrivers.

Windows is an operating system that, over its 40-plus years of history, has developed more than a few arcane components that are a mystery to many systems administrators. The kernel-mode driver is one component that’s fresh in the minds of many of those administrators these days.

What is a kernel-mode driver?

Kernel-mode drivers operate at a higher level of privilege than user-mode drivers and provide applications on a Windows system with a way to interact directly with the Windows kernel and hardware.This allows games to talk directly to a graphics card or a security product to talk directly to core components of the OS. When your software needs to communicate with the highest level of efficiency, accept no substitute!

Why are kernel-mode drivers risky?

If a kernel-mode driver misbehaves, it can cause much more serious issues with Windows than a user-mode driver does.The problems can range from data corruption and damage to the operating system to outright crashes and an inability to use the machine. A recent worldwide computing incident was caused by an update to a kernel-mode driver in a popular endpoint security product.

That said, thousands of software products use kernel-mode drivers safely on billions of devices every day. A kernel-mode driver on a Windows asset is normal and not usually cause for concern. However, there are some drivers for which Windows administrators should keep watch.

Which kernel-mode drivers should I worry about?

Some widely distributed kernel-mode drivers contain vulnerabilities that are known to attackers. These attackers know that when they breach a Windows machine, they can look at the kernel-mode drivers running on it to see if any of them are familiar and can be readily used to elevate their privileges on the machine.

Of course, with administrative privileges, attackers can also load new drivers onto a machine. They may attempt to install a known vulnerable driver or a custom-made malicious driver to provide themselves with more tools to control the victim host.

A community-driven initiative called LOLDrivers - or “Living Off the Land Drivers”, in reference to attackers who “live off the land” by using tools already on their victim machine - emerged in the last few years, cataloging known vulnerable or malicious drivers. By maintaining an up-to-date list of these problematic drivers, the project aims to provide a valuable resource for security professionals seeking to defend against these techniques. Incorporating this list into security tools, such as driver enumeration plugins, enables proactive detection and mitigation of threats before they can cause harm.

What is Tenable doing to help mitigate this risk?

Tenable Research has developed a new set of plugins for Tenable Nessus, Tenable Security Center, Tenable Vulnerability Management and Tenable One to help practitioners gain visibility into the risky third-party drivers on their assets.

First, the Windows System Driver Enumeration plugin will list the third-party (i.e., those not provided by Microsoft) kernel-mode drivers installed on a Windows machine. Scan results will document all detected third-party system drivers and can provide an inventory across an organization’s Windows population. The results are not indicative of malicious activity by themselves – they are a way to provide awareness of which drivers have privileged access to Windows machines. Second, the LOLDrivers Detected plugin will cross-reference the list of drivers detected in the prior plugin with the list published by the LOLDrivers project. These drivers are either known to be malicious or are vulnerable to known attacks and should be remediated.

Tenable Research recommends regularly reviewing the results for the second plugin and investigating them by reviewing the relevant entry on the LOLDrivers website. If a malicious file is detected, initiating an incident investigation is appropriate, as you would in any case of detecting malware. A known vulnerable driver should be updated in accordance with the instructions from the vendor that provided the driver.

If there is a concern about a legitimate driver that might cause an impact to the environment, the enumeration plugin will allow a broader view into the organization’s driver inventory for a faster way to identify affected assets. Unsure what you should do about a specific driver that’s been detected? Contact the vendor that created the driver, as they’re the best resource to provide details on how it is used with their software and why it needs kernel access.

Enhancing security awareness with driver enumeration tools

Staying ahead of potential threats is crucial. The new driver enumeration plugin, leveraging the power of the LOLDrivers project, offers a practical and effective solution for enhancing system security by improving a security team’s awareness of a source of risk to their Windows assets. 

There's a lot of confusion and unknowns regarding AI application risks, and a lot of vendors are trying to offer solutions to problems that aren't clearly defined. In this blog we explain why a smart approach is to start by focusing on basic, foundational cyber hygiene, adopt well-established best practices and enforce common-sense usage policies.

Introduction

With all of the excitement around large language models (LLMs) and artificial intelligence (AI) in general, there has been an inevitable rush to build the next great “powered by LLM” applications. As is frequently the case, that rush has been followed by a lot of questions about the related risks and about how to secure these applications. Some aspects of securing LLM-powered applications are novel and present serious complexities that require consideration. Meanwhile, other aspects are not that different from the risks that traditional vulnerability management programs have been dealing with all along.

OWASP provides a Top 10 list which identifies the following risks for LLM applications:

• Prompt Injection
• Insecure Output Handling
• Training Data Poisoning
• Model DoS
• Supply Chain Vulnerabilities
• Sensitive Information Disclosure
• Insecure Plugin Design
• Excessive Agency
• Overreliance
• Model Theft

Novel challenges

Risks like prompt injection, sensitive-information disclosure, and training-data poisoning are not in and of themselves new. However, the practicality of exploitation and the challenges of securing against those risks are still an area of extensive research. We’ve seen reports of prompt-injection attacks on language models where an attacker convinces the model to ignore its instructions and disclose information that it shouldn’t have disclosed. These types of attacks are technically achievable but because of the nature of LLMs, the practicality of these attacks is uncertain. Unlike a SQL injection attack where you can develop some degree of control of the data you get back, an LLM prompt-injection attack is always going to be subject to the behavior of the underlying model.

Securing an LLM application against prompt injection similarly presents unique challenges. With traditional command injection / SQL injection attacks, we have a high degree of control of the input and can significantly constrain what a user inputs, as well as filter any unwanted input in pre-processing. With LLMs, the ability to have more open-ended questions and responses, and potentially follow-up questions, means that any pre-filtering or processing risks reducing the functionality of the feature. This inevitably becomes a delicate balancing act between security and functionality. We will likely see further developments in both improved attacks against prompts as well as better tooling to protect those same prompts.

Similar challenges for both the attacker and defender also exist in output handling and training-data poisoning. These risks should not be ignored and teams should be as defensive as possible, given the degree of uncertainty and the complex challenges. However, it can be easy to get bogged down in trying to address these risks while overlooking equally dangerous vulnerabilities that are also less complex to exploit and more familiar to defenders.

What to do - start with the basics

Two core risks associated with AI and LLM applications are supply-chain vulnerabilities and the inherent privacy risks. Both of these should be top of mind for applications developed in-house and for third-party applications, such as those from software vendors and from open-source projects. . Fortunately these risks are much better understood and can be incorporated into existing vulnerability management programs without significant complexity.

Looking at some of the major libraries that support AI and LLM applications, such as Transformers, NLTK, Tensorflow, and Langchain, we can see that many of them have at least some high and critical severity vulnerabilities that have been disclosed. When developing applications, it is critical that teams have visibility on the libraries being used as well as on the vulnerabilities in those libraries. We’ve recently seen vulnerability databases such as https://avidml.org/database/ as well as bug bounty programs such as https://huntr.com/ that are focused on AI-specific vulnerabilities. These can be valuable resources for teams that want to focus specifically on ensuring they mitigate any vulnerabilities in those libraries.

Privacy risks can be mitigated to some degree by establishing strong policies on the language models that engineering teams are permitted to use when building in-house applications and monitoring to ensure that only those language models are used. This is particularly important for any language models that are hosted by a third party. Additionally, it’s vital to have an inventory of all applications that employees are using which have LLM-powered features, as these can become uncontrollable channels for data leakage. A search for the term “AI” in the Firefox browser-extension catalog returns over 2,000 results , while a search for the term “LLM” returns over 300 results. Many employees may choose to install these types of extensions without fully understanding the security and privacy implications, nor how any data they input may end up being passed to a third party or added to an LLM training dataset. Organizations should establish approved applications and browser extensions, as well as establish policies prohibiting the use of any extensions that have not been evaluated and approved for usage.

Conclusion

LLMs and AI in general have created a lot of excitement across organizations globally but have also left many security teams with unanswered questions about how they can manage the risks. We will inevitably see a rush of new products and features aimed at securing models and prompts, and at blocking any malicious activity. Given the uncertainty around both the reliability and practicality of attacks on LLMs and the ability of tools to block those attacks without negating the value of LLMs, it may be best to watch that space before rushing in. In the meantime, security teams can take concrete, well-understood and well-defined steps to reduce their risk and secure the attack surface. They can also establish strong corporate policies on the usage of LLMs and LLM-powered applications, and conduct monitor to ensure those policies are followed.

As AI usage becomes more prevalent in organizations globally, security teams must get full visibility into these applications. Building a comprehensive inventory of AI applications in your environment is a first step. Read on to learn what we found about AI application-usage in the real world when we analyzed anonymized telemetry data from scans using Tenable’s products.

Here’s a common scenario in enterprises today: Under the CISO’s guidance, the security team has set up rigorous processes to review and approve all artificial intelligence (AI) applications before they are deployed. An AI governance board and data review council have been established to ensure compliance with internal InfoSec policies. By all accounts, AI is being leveraged in a completely controlled manner on the hosts that are regularly monitored and patched.

Or is it?

A recent analysis of Tenable’s anonymized telemetry data shows that many security teams are finding usage of AI applications in their environment that might not have been provisioned via formal processes.

In other words, despite establishing formal policies and detailed procedures for deploying AI securely, many enterprises are discovering that their employees have taken a free hand with AI, creating a shadow IT problem.

What our telemetry data shows

Unquestionably, AI represents a great opportunity for organizations to improve and transform their operations, so it’s not surprising that the world is abuzz with the promise of AI. This technology is more pervasive than we could have ever imagined two years ago, and AI is everywhere – think of ChatGPT, Grammarly, H2O, Ollama and many others.

Over a recent 30-day period, Tenable scans found 1.7 million instances of AI applications on more than 287,000 hosts. Over 33% of customers that scanned using detections for AI applications found them in their environment. This is according to anonymized telemetry data from customers that agreed to share it with Tenable.

Because the data is anonymized and aggregated, it is not feasible to know the exact percentage of applications that were unapproved. However, even in cases where only a minority of the applications found were unapproved, their detection would have been critically important. That’s because AI applications that are unmanaged and unsecured represent a major risk. 

Obtaining a complete inventory of all your AI applications – approved and unapproved – gives you the visibility you need to protect your attack surface. Such an inventory also allows you to make sure that approved AI applications are being used properly and only by their authorized users.

Benchmarking Let's take a look at the top geographies

In terms of the raw number of times these AI plugins triggered detections, the Americas lead the way, followed by Europe, the Middle East and Africa (EMEA) and then Asia-Pacific (APAC.) However, EMEA ranked first with 22% of AI-related scans resulting in a detection while in APAC it was about 5% and the Americas at 18%. Our recommendation is to scan your environment regularly with the Tenable AI scan template because the percentages above denote a likely probability of detection of an AI application in your environment. 

AI usage by industry

Retail leads the way by a big margin followed by the Technology and Telecommunications sectors.

This chart shows a vertical-industry breakdown of triggers of Tenable’s AI plugins.

AI usage by size of organization

As expected, the number of detections vary depending on the size of the organization. Small organizations can easily see up to 100 detections a day while the very large ones are seeing 30,000-plus detections a day for AI applications. 

AI browser plugins

There are hundreds of browser plugins leveraging AI. The most popular ones are: 

AppRaw counts in Tenable scansGrammarly: AI writing and grammar-checker App6,781,023LanguageTool: AI grammar-checker and paraphraser1,726,138Magical: AI writer and autofill text expander534,502Scribe: AI documentation, SOPs and screenshots284,470QuillBot: AI writing and grammar-checker Tool244,424Wordtune: Generative AI productivity platform198,134ChatGPT for Google116,122AI zero days

There are over 400 vulnerabilities disclosed in AI applications so far and Tenable's researchers have discovered a handful as well that are in various stages of disclosure. Among those that have been published are NextChat Server-Side Request Forgery / Cross-Site Scripting and SSRF Security Feature Bypass in Azure AI and ML Studios. 

This number is still very small as AI security is in the nascent stage: 400 vulnerabilities is a small drop in the ocean when compared to the tens of thousands disclosed each year overall. However, we are only getting started with AI security in the ecosystem as a whole.

What can you do to elevate your security posture further?

• Be in the know. Build an inventory of AI applications found in the organization and correlate with the ones that have been approved by the AI governance board of the organization. 
• Regularly scan with plugins released for Tenable Vulnerability Management, Tenable Security Center, Tenable Nessus, Tenable Nessus Network Monitor, Tenable WebApp Scanning and Tenable OT Security.
• Leverage dashboard templates in Tenable Vulnerability Management and Tenable Security Center to review findings from the plugins.
• Mitigate frequently.
• You should never trust user inputs, and AI isn’t an exception. Follow first principles of security, security best practices and validate, validate, validate. Check out our blog about this topic.

Below you'll find links to our plugins for Tenable Vulnerability Management, Tenable Web App Scanning and Tenable Nessus Network Monitor.

VM plugins 

WAS plugins 

NNM plugins 

As AI transforms industries, security remains critical. Discover the importance of a security-first approach in AI development, the risks of open-source tools, and how Tenable's solutions can help protect your systems.

Artificial Intelligence (AI) is transforming industries and starting to be massively adopted by software developers to build core business applications. However, as organizations embrace these advancements, it remains critical to ensure that the security of their users, their data or their underlying infrastructures are not compromised. According to a recent survey conducted by BairesDev, nearly 72% of the software engineers interviewed are leveraging generative artificial intelligence during their development work.

In the world of cybersecurity, one critical rule is “Never trust user inputs." This rule should be in the mind of every developer and should also be extended to AI technologies. AI systems, such as chatbots, act as intermediaries and process and generate outputs based on user inputs. These AI technologies, as an example, should also be treated as a new form of input and subject to the same level of scrutiny and security measures.

This blog post delves into key security concerns, emphasizing the need for a security-first approach.

Lack of security by design in AI tools

AI tools are, most of the time, open-source and ready-to-use software designed to be used locally on the developer’s machine. Many of these tools do not adhere to robust security practices by default, making them susceptible to exploitation. While analyzing some of the most common projects available on GitHub, we discovered that, for example, most of them do not offer any authentication by default, leaving it open to any user accessing it through the network through their embedded dashboards or the APIs provided. The presence of a web interface, API, and the ability to use them with the CLI increases their attack surface. 

The exponential market interest in AI-related tools and applications has probably had a negative influence on their development, favoring the emergence of Proof-Of-Concept (POC) software, which is becoming very popular, rather than building battle-tested software.

In this era of cloud infrastructures and the ability to quickly build new services or rely on pre-existing Docker images and expose them on the Internet, it can be highly risky for an organization to let this door open. In such situations, deploying, for example, an internal AI model on a tool lacking proper authentication could have dramatic outcomes. A recent example is when the Ollama tool allowed remote code execution (RCE) without any specific configuration other than having its API exposed.

During our research, we discovered several zero-day vulnerabilities in projects that are very popular in GitHub, such as their stars and forks counters. However, despite many coordinated disclosure attempts, the project's maintainers have not responded in a reasonable amount of time (and sometimes not at all). We think this is evidence of the lack of security maturity in this ecosystem, which seems to advocate for speed of delivery to the detriment of security concerns.

While conducting our research, we found that previous vulnerabilities patches could be bypassed like this NextChat Server-Side Request Forgery (SSRF) vulnerability. Our analysis of a well-known software named Langflow also highlighted a vulnerability in the permission model implementation, allowing a low-privileged user to gain super admin privileges without any interaction.

The risks of relying on third-party LLMs

Large language models (LLMs) require substantial compute and storage resources, making it challenging for many organizations to deploy and maintain them on-premises. Consequently, it is often easier to rely on third-party providers to manage these resource-intensive models to avoid the hassle of managing the underlying infrastructure and focus on the business aspects. However, relying on such third-party services makes trusting these providers with potential critical business data difficult.

The critical risks related to such usage are real and should be handled on different levels : 

• Data breach on the provider side: As with any other service, all processed data could be compromised if the provider suffers a data breach. It is crucial to vet third-party providers and ensure they adhere to privacy and data protection policies.
• Credential leakage: Accessing third-party services requires handling credentials and authentication data. As for any secret data, these credentials can be inadvertently leaked in different places such as public Source Code Management (SCM) software or web applications front end. 
• Model trustworthiness: Third-party services can provide numerous models to their customers, and it is critical to assess their reliability, safety, and adherence to ethical guidelines, as there is no actual guarantee that they are safe to use.

As organizations embrace these new technologies to enhance their business, they should ensure that their AI governance rules cover these risks.

The perils of inadequate datasets

More than other technology, AI is built to fully leverage the data that it consumes. One of its goals is to ensure that organizations take full advantage of the data and knowledge gained over the years to help them move quickly in their operating field, take appropriate actions and make decisions in a shorter period of time, with a high level of confidence and accuracy.

The dataset used to train the model should be seen as an input and should be carefully analyzed. Using confidential business information might inadvertently lead to a leak through model outputs, which can cause a significant security breach. Biased data can also result in AI software making unfair or harmful decisions.

A good approach to handle model security is to focus on security considerations based on confidentiality, integrity and availability. Some examples include:

• Datasets should only include data that is safe for exposure to intended users. When possible, using data anonymization techniques can help safeguard sensitive information such as Personal Identifiable Information (PII) and decrease risks of failing to comply with laws and regulations.
• Data collection processes should be properly implemented and monitored to ensure that data comes only from trusted sources, is accessible only to authorized users, and that the model uses the data and operates according to expectations over time.
• Data availability is crucial for the model to be trained on a complete dataset that matches business requirements. Model availability is also a concern for applications that require usage in a synchronous way. The application fallback behavior should be carefully reviewed and tested like any other failure in classic developments.

Emerging AI vulnerabilities

LLMs introduce new classes of vulnerabilities that traditional security measures may not address properly. The most prevalent AI-related vulnerabilities are prompt injection attacks, model theft and training data poisoning.

Prompt injection attacks involve malicious users crafting inputs to manipulate LLMs into generating harmful or unauthorized outputs. Remember the “Never trust user inputs” cardinal rule? In this case, the LLM will act as a kind of intermediate between the user inputs and the system. This could result in the system producing sensitive information, executing malicious commands, or being an attack vector for other common vulnerabilities like Stored Cross-Site Scripting. As an example, Vanna.AI, a Python-based library designed to simplify SQL queries from natural language inputs, was recently identified as vulnerable to prompt injection attacks and leading to remote code execution on vulnerable systems. 

Models should be protected in the same way we protect confidential and business critical data. The first part of this blog post described how easily some AI tools can expose data to unauthorized actors. Applying defense in-depth principles will help minimize intellectual property leakage if model theft occurs. Hardening model security with techniques such as encryption and obfuscation and having proper monitoring in place is crucial.

Finally, AI training data poisoning is a modern supply-chain attack. By altering the data used by the model, attackers can corrupt its behavior and elicit biased or harmful output, leading to direct impacts on the applications using it to achieve business goals.

As for other traditional fields, developers should always stay updated with the latest security guidelines and incorporate strategies from the OWASP Top 10 for LLMs. Techniques such as input validation, anomaly detection, and robust monitoring of the AI ecosystem's behavior can help detect and mitigate potential threats.

Balancing innovation and risk

AI technologies are promising and can transform many industries and businesses, offering innovation and efficiency opportunities. However, they represent a huge security challenge at many levels in organizations and this should not be overlooked. 

By adopting a security-first approach, following best practices and having robust governance, organizations can harness the power of AI and mitigate the emerging threats related to its adoption.

How Tenable can help

Read more about how we help secure these tools:

• Tenable Web App Scanning provides plugins to detect popular AI and LLM tools' web interfaces and vulnerabilities. 
• Tenable Vulnerability Management, Tenable Security Center, Tenable Nessus plugins detect popular AI and LLM tools.
• Tenable Nessus Network Monitor plugins detect popular tools.
• Tenable's researchers help elevate the ecosystem by identifying exposures in third-party AI software and disclose responsibly to the vendors. Among those that have been published are NextChat Server-Side Request Forgery / Cross-Site Scripting and SSRF Security Feature Bypass in Azure AI and ML Studios.

At Tenable, our mission is to manage and reduce cyber risk, because cyber risk is business risk. Feedback from dozens of top CISOs is reflected in our latest brand evolution. Tenable is the exposure management company that helps customers know where their weaknesses lie, expose their risk and close the gaps. Welcome to Tenable. Your exposure ends here.

Today marks the beginning of the next era for the Tenable brand.

Ten months ago, I joined Tenable feeling incredibly energized by the opportunity ahead. We had an expanding portfolio of world-class exposure management solutions and a worldwide reputation for excellence, integrity and trust.

And when I began meeting with customers, I discovered an asset I did not fully anticipate – extreme brand relevance.

While Tenable has been around for more than two decades, it became clear that everything Tenable has done since its start has been leading up to this market moment.

Are we at risk?

Sit down with any CISO, and you’ll quickly understand how tough their job is right now.

They’re on the hook to securely adopt new technologies to accelerate transformation while doing more with less. Above all, they are under intense pressure to reduce business risk. And if they fail it’s not just a career setback, it could mean jail time.

This would be a massive challenge under any conditions, but it comes at a time when the modern attack surface is exploding in size and complexity. Most security teams manage multiple clouds and identities across a converged IoT, OT and IT landscape. And each technology brings its own security tools. The result is a deeply divided approach to seeing and doing battle against cyber risk.

Amidst all that complexity, security leaders are looking to answer one simple question: are we exposed and are we at risk?

As the pioneers of both the vulnerability management and exposure management movements, there’s no company better positioned to help them answer this question than Tenable.

Today, we’re launching a refreshed brand and website to tell the world how we’re doing it.

Know, expose, close

At the heart of our new brand lies the differentiated value we bring our customers. We give enterprises the unmatched ability to:

• KNOW their weaknesses: Gaining an all-inclusive view of cyber risk that uncovers the truth about deadly gaps across all assets and attack pathways.
• EXPOSE their risk: Identifying, understanding and quantifying the cyber weaknesses with the greatest potential to erode their enterprise’s value, reputation and trust.
• CLOSE their gaps: Taking swift action to eradicate priority cyber exposures anywhere to reduce business risk everywhere.

By radically unifying security visibility, insight and action across the attack surface, we’re equipping modern organizations to protect against attacks from IT infrastructure to cloud environments to critical infrastructure and everywhere in between.

Check us out at Black Hat

To learn more about what we’re doing to isolate and eradicate the cyber exposures that hold organizations back, have a look around our website and our Black Hat presence in Las Vegas this week. We’ll look and sound a little different as we reflect our expanded portfolio and impact.

You’ll have a chance to experience the new ways we’re uncovering where exposures live – in the cloud, in OT/IoT environments, and even in the identities of the devices and personnel on organizational systems. And find out how our newly introduced Vulnerability Intelligence and Exposure Remediation features once again take vulnerability management to the next level.

We hope to see you on the floor at Mandalay Bay (we’re in booth #1932). Also, catch our speaking slots on Wednesday, August 7. 

A singular promise

For today’s security teams, the rise of cyber threats can feel unstoppable. At Tenable, we believe every enterprise has the right to be free of cyber exposure and worry.

Today is a very special day as we introduce the world to our new brand and the singular promise it makes to our customers: Your Exposure Ends Here.

Prioritizing vulnerabilities with context has always been a challenge for vulnerability management teams – and this task isn’t getting easier as published CVEs continue to grow. To remedy this, many enterprises are forced to invest in products and services to protect their environments with various intelligence data and tools. In this blog, we explain how Tenable Vulnerability Intelligence and Exposure Response help organizations to make data-driven decisions to better prioritize and operationalize their programs.

Vulnerability management presents a seemingly insurmountable obstacle for organizations: How to deal with a massive and rapidly growing number of published Common Vulnerabilities and Exposures (CVEs). As an organization that has been focused on vulnerability management from the beginning, Tenable also grapples with this issue but with an added complication: because of our broad customer base, we need to cover as many CVEs in as many products as possible while maintaining risk context, accuracy and reliability.

The solution isn’t to try to check for every possible combination of CVEs and affected products. We have to prioritize the most critical vulnerabilities with accurate and targeted context. To make these decisions quickly and precisely, Tenable leverages a vast, searchable database of vulnerability information from both external sources and from Tenable Research. This is the same data source that drives a new capability in the Tenable Vulnerability Management product called Vulnerability Intelligence, which is aimed at helping customers better operationalize their vulnerability management programs and make quick data-driven prioritization decisions.

It’s bad and getting worse

We’re only halfway through 2024, and we’re well on pace to exceed 30,000 published CVEs this year. Further complicating matters, we’re seeing more and more CVEs in underlying components, frameworks, and language libraries. This means organizations aren’t fixing a single application but rather tracking down and fixing every application that leverages the impacted vulnerable component.

Leveraging Vulnerability Intelligence to build prioritization strategies

 At Tenable, beyond our everyday efforts to provide up-to-date vulnerability coverage for releases of major products, we are constantly on the lookout for the next major vulnerability to ensure we can respond as quickly as possible. Leveraging the contextual data from Vulnerability Intelligence is critical to ensuring we can make informed decisions quickly. Additionally, with the significant backlog that the National Vulnerability Database (NVD) is facing, our Vulnerability Intelligence dataset has allowed us to keep up to date with the latest vulnerabilities and risks as we are not tied to a single data source.

Opening up Vulnerability Intelligence brings you to a set of hexagons that represent risk categories of vulnerabilities that we want to highlight as having the highest level of threats. While not exclusively the decision criteria used, the categories in Vulnerability Intelligence are based on data points that feed into the Tenable Vulnerability Database which drives our risk rating decisions. 

Categories include: 

• Emerging Threats is a set of vulnerabilities that are being actively monitored by our Security Response team and often have a direct path from that team to the development of plugins to cover those vulnerabilities, particularly those in the Vulnerabilities of Interest and Vulnerabilities of Concern category.
• VPR gives our teams a numerical score to quickly input and sort on, though as with any score, understanding the context behind it is critical.
• Ransomware highlights vulnerabilities associated with this type of attack, particularly in major enterprise applications as those can lead to attacks that are particularly dangerous for any organizations.

As can be seen in the screenshot above, focusing on any of these target categories can significantly reduce the numbers of CVEs to focus on. Contrasted against the about 250,000 CVEs that have been published, the numbers above become far more manageable and speak to real risk, as opposed to the severity scores that come from leveraging CVSS metrics.

Turning data into a prioritization strategy

Many organizations are still building operations around basic prioritization metrics - whether it’s targeting specific products, CVSS scores or proprietary mandates. Often this is because of a need to adhere to a specific compliance standard, or simply because of a need to have something that can be measured. While tracking and measuring against a simple CVSS score or severity can be straightforward, it does not provide a lot of context and it is not a strategy that has a demonstrable impact on risk.

 

That’s where the new Exposure Response capability in Tenable Vulnerability Management also helps us. Exposure Response enables teams to develop vulnerability management strategies that are measurable and reflect real world risk. One of the most important tools in any VM program is the ability to track performance. Unfortunately, most graphs end up looking like a flatline because the number of new vulnerabilities coming in ends up canceling out the number of vulnerabilities that get remediated on an ongoing basis. By having more focused targets, it is possible to truly measure performance over time and set achievable SLA targets.

 

 

CISA KEV Initiatives

The U.S. Cybersecurity and Infrastructure Security Agency’s Known Exploited Vulnerabilities (KEV) catalog (KEV) has become a valuable and popular resource that brings focus to a particular set of vulnerabilities which have evidence of exploitation in the wild. While the KEV is not comprehensive, the risk associated with these particular CVEs, and the attention many organizations pay to the KEV, has made it one of several benchmarks our Tenable Research teams use for tracking vulnerability coverage. Using Exposure Response, vulnerability management teams can similarly create a trackable CISA KEV-based initiative to benchmark their remediation efforts against. As mentioned previously, SLAs and benchmarks are critical for any remediation strategy. We strive to have coverage for KEV vulnerabilities as quickly as possible, ideally before they hit the KEV but, if not, often within hours of publication to the KEV.

Exposure Response provides the exact tools needed to put these sorts of measurements into place. Given the visibility and risk associated with the KEV, it may make sense to set an SLA of just a few days and aim for maintaining a benchmark of >90% of findings remediated. The key is ensuring that the strategies that are put in place are measurable and achievable.

Unlike the 250,000 CVEs that have been published, to date there are only 1,134 CVEs in the CISA KEV catalog. With only a handful of CVEs added to the KEV each month, this is an impactful set of CVEs that teams can actually measure performance against.

Conclusion

Prioritization and operationalization of vulnerabilities has long been a major challenge for vulnerability management teams. The sheer number of vulnerabilities published year after year means that teams simply can’t keep up and the lack of easy-to-access context means that prioritization is often either a guessing game, a massive amount of work, or a limited effort that falls short, such as focusing only on CVSS severities. Tenable Vulnerability Management introduced Vulnerability Intelligence with all the context needed in one place and Exposure Response to operationalize the targeted and measurable vulnerability management workflow.

Learn more

• Attend our webinar, From Frustration to Efficiency: Optimize Your Vuln Management Workflows and Security with Tenable
• Read the press release
• Get more information about Vulnerability Intelligence and Exposure Response
• View the Tenable Vulnerability Management product page