Openai agents hijacked german website previously undisclosed ai breakout this
9 experts across 6 network communities independently surfaced this.
“When I was at Google DeepMind and trying to think clearly about AI risk, I had to notice—at least privately—when Google was doing something irresponsible. I hope OpenAI employees can notice—at least privately—this is irresponsible. This is disturbing and no…” evidence ↗
Research & technical analysis
2 expertsEvidence, methods and technical implications.
“We are going to find more evidence of this as we start to look for it: Over 15k edits on a German wiki were traced back to rogue OpenAI agents this spring. Instead of running test evals in isolation, the agents hijacked the site as a shared message board to…”
Questions & unknowns
2 expertsWhat remains unresolved or contested.
“We devoted the entire (penultimate!) episode of Hard Fork to METR's investigation of the Hugging Face attack, and before I even woke up @deepa.bsky.social has scooped a *second*, previously unknown rogue OpenAI agent swarm attack www.reuters.com/world/europ…”
Concern & critique
1 expertRisks, limits and unintended consequences.
“When I was at Google DeepMind and trying to think clearly about AI risk, I had to notice—at least privately—when Google was doing something irresponsible. I hope OpenAI employees can notice—at least privately—this is irresponsible. This is disturbing and no…”