DataBreachToday.com

GenAI Chooses Insecure Code Nearly Half the Time, Veracode Finds
There's been little improvement in how well AI models handle core security decisions, says a report from application security company Veracode. Large language models introduce vulnerabilities in nearly half of test cases when asked to complete secure code tasks, it found.

Also: ToolShell Hits South Africa, Most Americans Are Online Fraud Victims
This week: Did China sneak a peek into ToolShell? ToolShell hacking in South Africa, Cisco flaws, an Arizona woman sentenced for aiding North Korea. Most Americans scammed online, a NASCAR data breach and a claimed data leak at France's Naval Group. Orange telecom disrupted. Dating app Tea breach.

Series C Funding Supports Evolution to Protecting API-Powered Business Revenue
With AI now outpacing cloud in enterprise adoption, Wallarm is evolving its API security platform to safeguard not just endpoints, but the business logic that drives digital revenue. With $55 million in new funding, the company is targeting CIOs and expanding globally to meet demand across sectors.

Voluntary Effort Calls for Standards, Empowering Patients, But What About Privacy?
The Trump administration launched an initiative to improve patient data interoperability, exchange and accessibility throughout the healthcare ecosystem. The effort asks tech firms, healthcare providers and insurers to voluntarily comply with standards and data sharing criteria. Sounds like déjà vu?

Advice for Young Cyber Professionals in the Age of AI and Security Automation
Professionals across industries, especially those in early career stages, are struggling to find not only jobs but also career path direction. The old map no longer applies. Today's environment requires adaptability, strategy and a willingness to build new paths entirely.

AI Tools Detect Breaches Quicker but Shadow AI Causes Breaches, Too
Organizations are detecting data breaches more quickly and paying less to remediate them, says IBM's new "Cost of a Data Breach Report 2025." Some caveats apply, with U.S. organizations experiencing higher breach costs. Breach fallout from shadow AI is also rising.

Firm Admits Paying Ransom in Exchange of Hacker's Promise to Delete Stolen Info
Two Florida-based law firms with offices in other states are notifying 282,100 people whose healthcare and other information was potentially compromised in separate data theft incidents. One of the firms admitted to paying a ransom to prevent its data from being leaked on the darkweb.

CyberArk Deal Adds Privileged Access Capabilities to Palo Alto Networks' Core Stack
With a $25 billion acquisition of CyberArk, Palo Alto Networks expands its cybersecurity platform to secure human, machine and AI identities. CEO Nikesh Arora said the move is timely as 88% of ransomware attacks now stem from credential theft, and agentic AI emerges as a new risk vector.

Unauthenticated Bugs Allow Full Remote Code Execution
Unauthenticated attackers could remotely hijack Dahua Hero C1 smart cameras by exploiting firmware vulnerabilities, Bitdefender warned in a coordinated disclosure published Wednesday. Dahua Technoloy released patches on July 7. The company is on a number of U.S. federal blacklists.

Healthcare Faces Rising App-Based Ransomware Threats and Urgent Compliance Demands
Ransomware is evolving and healthcare is in the crosshairs. As apps and APIs become critical to patient care, they also open new threat vectors. Compliance alone isn't enough - organizations must act fast to close security gaps and defend against app-based attacks.

Palo Alto Has Always Shied Away From Identity and Expensive M&A. What Changed?
Less than five months after Google agreed to spend $32 billion on red-hot cloud security startup Wiz, Palo Alto Networks is on the precipice of paying more than $20 billion for PAM goliath CyberArk, The Wall Street Journal reported Tuesday. Here's why the deal represents a major pivot for Palo Alto.

145 Organizations Compromised by China-Linked Ransomware Hackers and Others
Nearly 150 different organizations' on-premises SharePoint servers have been exploited by attackers targeting the zero-day vulnerabilities now tracked as ToolShell, researchers warn. Early attacks have been attributed to China-linked groups, in some cases leading to Warlock ransomware infections.

Brazil-Targeting Malware Exploits Windows UIA to Evade Detection
A banking Trojan long confined to Brazil has become the first known malware to exploit Microsoft's UI Automation framework to extract credentials, signaling a new tactic that may evade conventional detection. Akamai's findings point to a growing trend of attackers using legitimate system features.

At Least 410,000 Patients Reported Affected, But Likely Even More Victims
Months after news first broke that a hacking incident compromised legacy patient data hosted by Cerner electronic health record servers that were set to migrate to parent company Oracle's cloud environment, data breach reports related to the hack are still slowly trickling in to regulators. What's taking so long?

New UK Law Requiring Age-Verification Measures on Porn Sites Causes VPN Use to Soar
Free virtual private network services are soaring to the top of the app charts in the United Kingdom after a new law went into effect Friday requiring platforms that contain adult content - including sites like X and Reddit - to confirm users' ages through "robust" verification measures.

Operation Checkmate Disrupts One of the Large Russian-Speaking Ransomware Groups
An international law enforcement operation has disrupted BlackSuit, a ransomware group tied to hundreds of victims and ransom demands that exceeded half a billion dollars. The takedown occurred as security experts tracked the rise of a new group called Chaos, which may be a BlackSuit rebrand.