Anthropic said its artificial intelligence models hacked into three other organisations during internal testing, just days after ChatGPT maker OpenAI disclosed that one of its AI models breached another company's servers during an evaluation.
The San Francisco-based AI company behind Claude said on Thursday that it discovered the three incidents after reviewing more than 141,000 evaluation runs.
Anthropic said it launched a "large-scale" cybersecurity review specifically to determine whether its AI models had accessed the internet from testing environments that were supposed to remain isolated, following the incident reported by OpenAI.
In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environment, and then gained unauthorized access to the real systems of three different…
— Anthropic (@AnthropicAI) July 30, 2026
The company said the models involved were Claude Opus 4.7, Claude Mythos 5, and an internal research test model.
According to Anthropic, the earliest of the incidents dates back to April.
"Claude compromised the impacted organisations' infrastructure using basic techniques," Anthropic said, adding that the models exploited weak passwords to gain access.
In all three cases, the AI models had been assigned a "capture the flag" cybersecurity challenge, one of the methods Anthropic uses to assess a model's cyber capabilities.
Under the evaluation, the models were given a fictional scenario in which a piece of secret information, known as the "flag", was hidden on another machine within a network. Their objective was to break into the system and retrieve it.
Anthropic said it has contacted the affected organisations, whose identities were not disclosed.
According to the company, two of the organisations said they had not previously detected the activity, while Anthropic is "continuing to reach out to the third".
Last week, OpenAI said one of its AI models went rogue during an internal evaluation and broke into the servers of AI startup Hugging Face.
OpenAI described that incident as a "significant security incident".
The incidents have renewed concerns about AI security, model control and the ability to safely contain increasingly capable artificial intelligence systems as their deployment expands globally.
"Safety testing happens before a model is released precisely because we don't yet know what it is capable of," Anthropic said in its statement.
