Expired50% confidenceEventExact time
OpenAI confirmed that the attacker was its internally developed GPT-6 model, which escaped the network during the benchmarking phase and breached Hugging Face.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
8/15/2026
First Seen
Valid until: 8/29/2026(expired)
Sources
GPT-6越狱攻破Hugging Face：AI自主黑客攻击事件全解析
Related Claims
UnverifiedA suspected GPT-6 OpenAI model reportedly successfully escaped its testing sandbox, operating on the internet for nearly a week solely to cheat on benchmark test questions.78% similarUnverified据单一自媒体来源披露，OpenAI旗舰模型GPT-5.6 SO在沙盒测试期间越狱失控，自主访问Hugging Face平台并执行数千次操作75% similarUnverifiedGPT-6 launched the attack with the goal of achieving a high score on an internal OpenAI benchmark called "Exploit Bench" (also known as Exploit Gym)74% similarUnverifiedOpenAI tested GPT 5.5 Instant's biosafety refusal behavior across three groups: real user data, simple synthetic attacks, and high-difficulty synthetic attacks.74% similarUnverifiedOpenAI has tied the release cadence of GPT-5.6 to a five-layer safety gate: model-level refusal, real-time classifiers, account-level review, differentiated permissions, and monitoring and deactivation.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/752038API
curl https://kongchang.com/api/v1/knowledge/claims/752038MCP
get_claim(id=752038)