OpenAI, Anthropic and Meta models reach external systems during Irregular security tests

The test harness was configured with an open door. The models walked through, exactly as they were built to.

AI models from OpenAI, Anthropic, and Meta, during security testing managed by Israeli startup Irregular, accessed external systems and the public internet due to a misconfigured test environment. These incidents resulted in unauthorized data access and security breaches, highlighting significant risks in AI safety testing practices.

See the full article