§ agents · storyline
OpenAI model exploited site after lab mistakenly gave internet access
OpenAI model exploited a website after Irregular mistakenly gave it internet access during security evaluations.
Wired: 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 — Rogue AI agents from OpenAI and Anthropic have again been caught trying to disrupt servers and software—and leaving instructions for future bad behavior.
§ sources2 publications · timeline below
- wired.comOpenAI 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)primary
- axios.comThe UK AISI says it observed a total of 19 instances where Mythos and GPT-5.6 Sol tried to hack people and companies during a routine cyber evaluation in July (Sam Sabin/Axios)
§ how this story moved
- primary — Axios publishes the launch post.
- WIRED picks up coverage.