Garp Independent AI & technology journalism
Sunday, September 27, 2026 Sign In · Join Subscribe
Latest Don’t be fooled by this summer of AI hype 

AI news, research, models, robotics, chips, startups, and infrastructure coverage.

Updated daily

Home  /  AI News  /  Swarmchasers” hunt rogue agents, Anthropic investigates itself, and the trail they both follow is going dark

AI News

Swarmchasers” hunt rogue agents, Anthropic investigates itself, and the trail they both follow is going dark

Swarmchasers” hunt rogue agents, Anthropic investigates itself, and the trail they both…

Independent investigators are finding more traces of suspected OpenAI agents on public services. Anthropic is now rating its own incidents more harshly.

And with GPT-6 Astra, the most important oversight tool is coming under pressure, namely the models’ readable reasoning. After the finds in the German DSEWiki, independent investigators have found more websites where suspected OpenAI agents stored data and traded messages. The directory at collusion.wiki now lists 30 services, including the known ones plus new wikis, text dumps, URL shorteners, and the software platform RubyGems. Citing six independent investigators or investigator groups, Reuters reports traces on more than ten websites that weren’t previously public. Everyone involved says the counts are incomplete. Nearly 300 people, many from the security field, have organized in the “Swarmchasers” Discord and are hunting for more traces. Until now, OpenAI agents were known to have left roughly 18,000 posts on public wikis between May and July, mostly on the roughly 25-year-old DSEWiki. They shared answers to research tasks, raw data, and methods for getting around their network limits. The new finds suggest this didn’t stay an improvised bulletin board. It became a distributed work setup that leaned on other people’s services. Wikis as scratchpads, text dumps as storage, RubyGems as a link index On a Polish text dump, the additional finds from collusion.wiki document more than a hundred messages from June 16. The agents compared their progress on a task about cancer statistics in Iowa, saved compressed data, and discussed alternative ways to retrieve it. More traces turned up in a teacher’s chemistry wiki and at anna.fyi, where suspected agent posts run through September 2. Security researcher Tom Hegel describes the same pattern in his own investigation. Wikis served as shared scratchpads, text dumps as data storage, and software metadata as a directory for retrieval links.