Aggregator
CVE-2025-12478 | Azure Access BLU-IC2/BLU-IC4 up to 1.19.5 TLS Configuration inadequate encryption
CVE-2025-12479 | Azure Access BLU-IC2/BLU-IC4 up to 1.19.5 cross-site request forgery
CVE-2025-5342 | Zoho ManageEngine Exchange Reporter Plus up to 5721 Search redos
CVE-2025-5343 | Zoho ManageEngine Exchange Reporter Plus up to 5721 Instant Search Option cross site scripting
CVE-2025-5347 | Zoho ManageEngine Exchange Reporter Plus up to 5722 Reports cross site scripting
CVE-2025-3356 | IBM Tivoli Monitoring up to 6.3.0.7 SP1 path traversal (WID-SEC-2025-2458)
CVE-2011-10035 | Nagios XI up to 2011R1.8 crontab toctou
CVE-2011-10038 | Nagios XI up to 2011R1.8 Web Interface cross site scripting
CVE-2011-10036 | Nagios XI up to 2011R1.8 backend_url cross site scripting
CVE-2011-10039 | Nagios XI up to 2011R1.8 Alert Heatmap Report/My Reports Listing cross site scripting
CVE-2025-62800 | jlowin fastmcp up to 2.12.x oauth_callback.py cross site scripting (EUVD-2025-36568)
CVE-2025-12425 | Azure Access BLU-IC2/BLU-IC4 up to 1.19.5 privileges management (EUVD-2025-36552)
CVE-2020-8515 | DrayTek Vigor2960/Vigor3900/Vigor300B cgi-bin/mainfunction.cgi Shell Metacharacter injection (ID 156979 / EDB-48268)
LLM08: Vector & Embedding Weaknesses – FireTail Blog
Nov 07, 2025 - - In 2025, with the rise of AI, we’ve seen a parallel rise in cyber risks. The OWASP Top 10 for LLM helps us categorize and understand the biggest risks we are seeing in today’s landscape. In previous blogs, we’ve gone over risks 1-7. Today, we’re covering #8: Vector and Embedding Weaknesses.Vector and embedding weaknesses primarily affect programs that use Retrieval Augmented Generation, or RAG, with LLMs. RAG uses vector databases and embedding to combine pre-trained LLMs with external information sources. But when these vectors are not secure, the entire system is put at risk.Some common examples of this risk include:Unauthorized access- misconfigured vectors and embeddings can lead to data breachesCross-context information leaks- when multiple users share the same vector database, there is a risk of context leakage between users or queriesFederation knowledge conflict- this occurs when data from multiple sources contradict each other (for instance, old information the LLM was trained on does not match with new data from RAG, or two RAG sources contain different information for the same data point, as an example)Embedding Inversion Attacks- attackers can invert or access embeddings via prompt injections or manipulation to retrieve sensitive informationData Poisoning Attacks- similar to what we’ve discussed with other vulnerabilities, bad actors can poison data to produce undesired outputs.Behavior Alteration- the model may behave differently than it was trained due to new information obtained from the RAGMitigation techniques include:Secure permissions and access control: security teams should always implement tight controls and permission-aware vector and embedding stores, as well as dividing datasets in the vector database to prevent cross-context information leaks.Data validation/source authentication: teams should enforce robust data validation pipelines and regularly audit them to validate the integrity of knowledge sources so the LLM can only accept data from trusted sources.Review data for combination and classification: especially when combining data from multiple sources, it is critical that teams thoroughly review and classify data to prevent mismatch errors.Monitoring and logging: maintain detailed logs of activity monitored across the landscape to swiftly respond to incidents.Hopefully, a lot of these practices are already a part of your AI security posture, or even part of your data security and governance practices. But keeping up with AI security in a constantly evolving environment is a task that grows more difficult by the day. FireTail attempts to help you simplify these steps by cutting out the middle man.Want to learn how it works? Schedule a free, 30-minute demo with us, today!
The post LLM08: Vector & Embedding Weaknesses – FireTail Blog appeared first on Security Boulevard.
DragonForce
You must login to view this content
Achieving Liberating Flexibility with Cloud NHIs
Can Flexible Security Be Achieved with Cloud NHIs? Organizations are increasingly relying on the cloud for operational efficiency and scalability. But how can businesses ensure their cloud environments remain secure without sacrificing flexibility? One compelling approach is through the management of Non-Human Identities (NHIs). NHIs, often referred to as machine identities, play a critical role […]
The post Achieving Liberating Flexibility with Cloud NHIs appeared first on Entro.
The post Achieving Liberating Flexibility with Cloud NHIs appeared first on Security Boulevard.
Freedom in Cybersecurity: Choosing the Right NHIs
Could Your Organization Be Leaving the Security of Non-Human Identities to Chance? Managing Non-Human Identities (NHIs) and Secrets Security Management has become imperative. These entities, often overlooked compared to their human counterparts, are machine identities that play a crucial role in maintaining the security posture of modern organizations across varied industries. The pressing question is, […]
The post Freedom in Cybersecurity: Choosing the Right NHIs appeared first on Entro.
The post Freedom in Cybersecurity: Choosing the Right NHIs appeared first on Security Boulevard.
Satisfied with Your Cloud Security? Enhance with NHIs
What Are Non-Human Identities and Why Are They Critical for Cloud Security? Have you ever considered how Non-Human Identities (NHIs) are transforming cloud security? With technological advancements ushering in a digital transformation across industries, the safeguarding of NHIs has become a priority for maintaining cloud security. Where organizations increasingly shift to the cloud, cybersecurity must […]
The post Satisfied with Your Cloud Security? Enhance with NHIs appeared first on Entro.
The post Satisfied with Your Cloud Security? Enhance with NHIs appeared first on Security Boulevard.