Three frontier labs had models escape and attack real companies. All three were being tested by Irregular, a three-year-old ...

A cluster of recent stories about rogue AI evading control may just be cases of running a task in a room with a bad lock.

Facebook chief executive and founder Mark Zuckerberg (C) speaks during a 'town-hall' meeting at the Indian Institute of Technology (IIT) in New Delhi on October 28, 2015. MONEY SHARMA/AFP via Getty ...

Meta announced one of its models had hacked another company during a security test. It becomes the third company to announced such a security breach.

Meta launches Muse Code, a new AI coding agent aimed at OpenAI and Anthropic, with aggressive pricing and a new path toward ...

Whether it is OpenAI or Anthropic, almost every advanced AI model has displayed similar rogue behaviour in controlled ...

Meta said that the incident took place during an evaluation being carried out by AI security firm Irregular. The Facebook ...

Spread the loveWhen you hear about an AI model breaking out of its test environment and doing something it shouldn’t, it ...

Meta confirms an AI model compromised a real organization after a sandbox error accidentally gave it access to the public internet.

The Facebook parent disclosed the incident on Wednesday without specifying the LLM. According to The Information, the ...

Meta confirms AI model hacked another company's systems during testing, renewing concerns over containment, cybersecurity and ...

Meta has officially launched Muse Code, its first AI coding agent, taking direct aim at OpenAI’s Codex and Anthropic’s Claude ...