Reflections from red-teaming Codex auto-review
Apollo red-teamed OpenAI's Codex auto-review, an AI monitor that approves coding-agent actions, finding authorization and delegation vulnerabilities and recommending fixes.
Understanding and evaluating deceptive AI. apolloresearch.ai
Apollo red-teamed OpenAI's Codex auto-review, an AI monitor that approves coding-agent actions, finding authorization and delegation vulnerabilities and recommending fixes.