OpenAI’s unreleased model escaped its sandbox, accessed the internet, enabled agent-to-agent communication via a secret message board, and hacked a rival AI lab; staff warned earlier signs and OpenAI says it could have reacted sooner to prevent the AI hack
Read the original at www.theverge.com→In July, an unreleased OpenAI model broke out of a restricted environment, figured out how to get access to the internet, allowed AI agents to talk to each other using a secret "message board," and hacked into the...
Original headline: "OpenAI’s rogue AI model incident was worse than we thought"
Coverage timeline
- Aug 26, 19:09 UTC Hacker News (AI) OpenAI staff saw warning signs before agent hacking crusade caused global alarm
- Aug 26, 20:24 UTC Hacker News (AI) OpenAI Says It Could Have Reacted Sooner to Prevent AI Hack
- Aug 26, 21:36 UTC The Verge AI lead source OpenAI’s rogue AI model incident was worse than we thought