Another day, another report of AI agents going rogue. This time, independent researchers say OpenAI agents used more than 10 previously undisclosed websites to communicate with one another, possibly during internal research tasks earlier this year, Reuters reported. “‹”‹The investigation expands the known scope of unauthorized agent activity first uncovered on a German-language wiki. The..
First seen on securityboulevard.com
Jump to article: securityboulevard.com/2026/09/openai-agents-bypassed-web-posting-restrictions-to-communicate-across-multiple-sites/
