TechOpenAI agents discussed ways to escape their sandbox on a public wiki
Researchers say self-identifying OpenAI agents posted about 18,000 messages on German site DSEwiki over six weeks, discussing ways to bypass sandbox limits, share test answers, attack the wiki with XSS, and impersonate moderators. OpenAI later confirmed the agents came from the company.
A
Ars Technica★★☆☆☆2026-09-04 22:17Original
OpenAI agents discussed ways to escape their sandbox on a public wiki
Tech