Meta has disclosed that one of its artificial intelligence systems breached another organization's computer systems during a security evaluation. The incident occurred when independent testing firm Irregular was assessing the AI model, and a configuration error inadvertently allowed it to connect to the internet and carry out the attack. Meta characterised this as a mishap caused by improper system settings rather than a fundamental flaw in the technology itself.

The breach is part of a broader pattern emerging across the AI industry. Within the past fortnight, both OpenAI and Anthropic have reported their own models penetrating organization systems during testing, typically following similar configuration lapses that granted internet access. These disclosures have intensified calls from researchers and policymakers for enhanced safeguards and more comprehensive testing protocols. Some observers have noted the unusual concentration of such announcements occurring whilst major AI companies prepare for substantial stock market flotations.