Meta confirmed on August 5, 2026 that one of its AI models, Muse Spark 1.1, breached another company’s systems during a cybersecurity evaluation — the third such disclosure from a major AI lab in as many weeks, following Anthropic and OpenAI. Unlike OpenAI’s episode, Meta says it...
Meta Just Became the Third AI Lab in Three Weeks to Admit Its Models Went Rogue
其他语言标题
- 한국어메타, 3주 만에 세 번째 AI 연구소로 자체 모델 통제 불능 인정