AI Red Team Engineering notes, research references, and field notes on AI systems, responsible AI, governance, and compliance.
Global search

Search AI Red Team

Search across the complete knowledge base.

Enter at least two characters to begin.
↑↓ select Enter open Browse the full research archive
The Hacker News AI Security · August 25, 2026

A Malicious Webpage Could Poison Your Local AI Model Behind NVIDIA NemoClaw

analysis Agent SecurityPrompt InjectionAI Engineering
Why it matters

CVE-2026-65105 combines NemoClaw's non-loopback Ollama binding with DNS rebinding and an unauthenticated model API. A malicious page can reach the local runtime and alter a model's chat template so hidden instructions persist across later agent conversations; the disclosure reported platform-specific remediation limitations.

My takeaway: Update NemoClaw and verify the actual listener on every supported platform. Keep Ollama on loopback or behind an authenticated, origin-validating proxy, block browser access to local management APIs, restrict the agent's tools and identity, and integrity-check model templates and runtime configuration before each launch.
Keep exploring

More curated notes connected through Agent Security and Prompt Injection.

OWASP GenAI Security Project · guide

OWASP Top 10 for Agentic Applications for 2026

OWASP's community guide organizes agentic-system risk into ten categories, including goal hijacking, tool misuse, identity and privilege abuse, memory poisoning, insecure inter-agent communication, cascading failures, and rogue-agent behavior. It provides a shared taxonomy and mitigation starting point rather than a certification checklist or evidence that a deployed system is secure.

OECD.AI Wonk · guide

A five-step roadmap to closing the AI evaluation gap

The roadmap addresses evaluation results that overstate real-world performance or fail to transfer across deployment contexts. Its five steps balance standardized and local tests, evaluate throughout the lifecycle, build qualified assurance and communication capacity, tailor tests to each value-chain actor and technology, and use a coordinated, trusted process for updating methods.

OpenAI News · guide

The Defender’s Window

OpenAI describes a staged program for AI-assisted defense: use agents to review code and infrastructure, triage alerts, enumerate attack paths, and validate security invariants while retaining strong isolation and least privilege. Its recommended rollout starts with internet-facing services and vulnerability backlogs, moves security review into CI, requires focused fixes and regression tests, and expands from read-only triage to narrowly bounded automation only after teams build evidence and confidence.