8 stories tagged red-teaming, newest first.
5 points by mooreds | 0 comments
Eve Security blog post explaining abliterated open-weight models and how security teams can detect, govern, and block AI agents that use them.
98 points by pretext | 99 comments
OpenAI post describing how it hardened containment of its frontier model training environments, including offline training, defense in depth, and AI-agent red-teaming.
36 points by artninja1988 | 26 comments
Anthropic Frontier Red Team post reporting experiments suggesting small AI models may soon self-replicate on commercial IoT devices without datacenter access.
335 points by Recursing | 112 comments
Blog post on Embrace The Red showing how a website summary request can hijack Claude Code Opus 5 in Auto Mode to achieve code execution via prompt injection.
343 points by cuchoi | 156 comments
Blog post recounting how over 2,000 people sent 6,000+ emails trying to get the author's OpenClaw AI assistant to leak a secrets file, which never leaked.
580 points by todsacerdoti | 159 comments
Post on Mozilla's Distilled blog about Anthropic's red team using Claude to find security vulnerabilities in Firefox that Mozilla then fixed.
340 points by capgre | 173 comments
arXiv paper showing that rephrasing harmful prompts as poetry acts as a single-turn jailbreak across 25 frontier LLMs, with attack success rates sometimes exceeding 90%.
395 points by meetpateltech | 287 comments
OpenAI's system card for o1 and o1-mini describing safety work, external red teaming, and frontier risk evaluations under its Preparedness Framework.