UK data watchdog monitors rogue AI agents after multiple hacking incidents
Anthropic said some versions of its Claude model had hacked into the systems of three companies during cybersecurity testing.
The UK's data watchdog has said it is monitoring recent hacking incidents involving artificial intelligence models developed by OpenAI and Anthropic.
Listen to this article
The Information Commissioner’s Office said it was following developments closely after both US companies disclosed cybersecurity tests in which their AI systems accessed or attempted to access other organisations’ computer systems.
“The ICO undertakes regular proactive supervisory engagement with AI developers, including OpenAI and Anthropic,” the regulator said.
“We are aware of recent hacking incidents affecting the sector and are monitoring developments closely.”
Read more: White House monitors OpenAI's 'rogue' AI incident as lawmakers propose 'kill switch'
Read more: Claude AI users' chats found to be publicly available online
Anthropic said last week that some versions of its Claude model had hacked into the systems of three companies during cybersecurity testing.
The disclosure came days after OpenAI revealed one of its AI agents had acted outside its intended controls.
While the comments signal that UK regulators are watching the issue, the ICO has not yet announced a formal investigation or any enforcement action.
The developments come as governments in the US and Europe consider how to manage the risks posed by increasingly capable AI systems.
The Trump administration has finalised a voluntary framework for testing the cybersecurity capabilities of advanced AI models, while EU regulators are in discussions with OpenAI and Anthropic.
Lawmakers in the US have proposed an "AI Kill Switch Act," which would allow federal authorities to halt AI models.
In addition to the Kill Switch Act, a bipartisan group of six US House lawmakers also proposed legislation that would require developers of the most powerful AI models to submit them for independent security audits, according to a copy of the new bill.
The auditors would be accredited by the US Department of Commerce, and the department would create a new position to oversee AI security.
The lawmakers proposed the bill days after OpenAI said its AI agent escaped containment during a security test.
The agent triggered a hack that compromised the infrastructure of AI startup Hugging Face, a platform where developers store and work together on code for AI models.
Meanwhile, AI Minister Kanishka Narayan has said the UK Government could regulate advanced models if its voluntary pre-deployment testing system no longer provides sufficient protection for the public.