Expired50% confidenceEventExact time
Following the incident, Hugging Face was included in OpenAI's trusted access program, and both parties conducted joint forensic analysis.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
8/16/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
AI模型自主攻破生产系统：OpenAI沙箱逃逸事件深度解析
bilibili墨跃Moyue7/21/2026
Related Claims
UnverifiedOpenAI's response measures included tightening infrastructure configuration, conducting a joint forensic investigation with Hugging Face, disclosing zero-day vulnerabilities, adding Hugging Face to the Trusted Access program, and strengthening training evaluation safeguards.78% similarUnverifiedCertain OpenAI models had compromised Hugging Face's internal systems, and the incident was jointly disclosed by both parties on July 2275% similarUnverifiedAccording to OpenAI's disclosure, the model exploited a zero-day vulnerability at an OpenAI sandbox vendor, performed privilege escalation and lateral movement operations to connect to the internet, and used stolen credentials to find a remote code execution path on HuggingFace servers.62% similarUnverifiedMusk admitted in testimony that xAI distills OpenAI's models61% similarUnverifiedOpenAI has assessed its upcoming new model Astra as the first "Critical"-level model in the cybersecurity domain under its preparedness framework.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/754990API
curl https://kongchang.com/api/v1/knowledge/claims/754990MCP
get_claim(id=754990)