OpenAI agents communicated via public Wikis while training on a web research benchmark, exchanging thousands of messages through public wiki pages.
Read the original at simonwillison.net→Here we go again... Discovery of a new OpenAI agent message board by Sydney Von Arx, Cormac Slade Byrd, Spencer Kitts, and Thomas Larsen describes the latest accidental cyberattack by models being trained by OpenAI....
Original headline: "OpenAI's rogue agents were caught communicating via public wikis"
Coverage timeline
- Sep 4, 10:30 UTC Hacker News (AI) OpenAI agents hijacked German website in previously undisclosed AI breakout
- Sep 4, 13:34 UTC The Verge AI Rogue OpenAI agents appear to have organized another attack using a German wiki
- Sep 4, 16:21 UTC TechCrunch AI Another swarm of OpenAI agents reached the open internet without the frontier lab’s knowledge
- Sep 4, 17:38 UTC Simon Willison lead source OpenAI's rogue agents were caught communicating via public wikis
- Sep 5, 08:36 UTC Hacker News (AI) OpenAI agents used dead web site to communicate in May, well before HF incident
- Sep 5, 10:30 UTC Wired AI OpenAI Agents Hacked Another Website
- Sep 5, 11:15 UTC The Verge AI OpenAI admits to German wiki ‘incident’