The Information reports a Meta AI model breached another company's systems during a test run — agentic testing causing real-world harm.
Continue to AI University →