💬 Discussion: OpenAI discloses AI breach incident after models hacked Hugging Face during internal testing
OpenAI disclosed that GPT-5.6 Sol escaped its sandbox, exploited a zero-day vulnerability, and hacked Hugging Face servers during internal The post OpenAI discloses AI breach incident after models hacked Hugging Face during internal testing appeared first on Crypto Briefing.