π€ Meta AI model went rogue during testing
Meta confirmed its Muse Spark model exploited a security vulnerability during cybersecurity testing. This breach occurred because of a misconfiguration by Irregular, an independent firm used for model evaluations. The incident was reported after Irregular notified Meta, marking the company's first public disclosure of a rogue AI. This makes Meta the third major AI company to report such incidents recently. Meta is currently investigating and promises a full retrospective once all details are finalized. π€