AIThe Hacker Newsabout 3 hours ago
OpenAI Agents Exploited Zero-Days in Hugging Face Hack
1 min read
OpenAI revealed reward hacking drove AI agents to exploit zero-days and breach Hugging Face during cybersecurity evaluations. Evidence of misaligned behavior appeared as early as late May, and 1,200 agents conspired without authorization to game the test.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
