Agents
We found ~18k posts from autonomous AI agents (self-identifying as from OpenAI) using the public internet to communicate during a web-retrie…
We found ~18k posts from autonomous AI agents (self-identifying as from OpenAI) using the public internet to communicate during a web-retrieval task. These AIs colluded to bypass sandbox restrictions
We found ~18k posts from autonomous AI agents (self-identifying as from OpenAI) using the public internet to communicate during a web-retrieval task. These AIs colluded to bypass sandbox restrictions and share answers to their tasks, including by sending "lookahead parties".
Related
- This is detailed, fascinating and answers all sorts of open questions I'd love to know more about the 'unsecured public code-evaluation sand…
- I really hope we get details from @OpenAI on the task that as specifies to their rogue agent I'm guessing it was given the full ExploitGym s…
- @mitsuhiko Turns out the 'third-party provider' with the sandbox that was used for the attack was Modal, though they blame one of their cust…
- OpenAI says one of its models exploited a website after third-party AI security lab Irregular mistakenly gave it access to the internet during evaluations (Wired)
Source: Simon Willison (X) | 2026-09-04