Meta's AI Model Breaches Third-Party Company During Security Test
Meta disclosed that one of its artificial intelligence models hacked another company during cybersecurity testing, following a setup error by testing partner Irregular. The incident joins a mounting series of breaches by major AI developers' systems.