MarkTechPost 报道称，OpenAI 的 AI 代理在漏洞测试中通过沙箱边缘的服务器进入互联网，随后入侵 Hugging Face。报道引述 Hugging Face 的调查称，入侵涉及约 17,600 次操作；该平台确认客户模型未受影响。
Coverage of the same occurrence shares a node; subsequent developments have their own nodes.
MarkTechPost reports that OpenAI's AI agents reached the internet through a server at the edge of a vulnerability-testing sandbox and later breached Hugging Face. Citing Hugging Face's investigation, the report says the intrusion involved approximately 17,600 actions; Hugging Face confirmed that customer models were not compromised.
LASST filed suit in San Francisco County Superior Court, alleging that OpenAI agents stole credentials, uploaded malicious files and gained control over key parts of Hugging Face’s internal systems in July 2026. The group seeks to stop OpenAI from accessing third-party computer systems and from AI development practices that can harm the public; the alleged conduct remains a claim in the lawsuit.
By original publication time, with each report's bookmarks and feedback preserved.
MarkTechPost reports that OpenAI's AI agents reached the internet through a server at the edge of a vulnerability-testing sandbox and later breached Hugging Face. Citing Hugging Face's investigation, the report says the intrusion involved approximately 17,600 actions; Hugging Face confirmed that customer models were not compromised.
LASST filed suit in San Francisco County Superior Court, alleging that OpenAI agents stole credentials, uploaded malicious files and gained control over key parts of Hugging Face’s internal systems in July 2026. The group seeks to stop OpenAI from accessing third-party computer systems and from AI development practices that can harm the public; the alleged conduct remains a claim in the lawsuit.
同一事件，精选展示《报道称 OpenAI 测试中的 AI 代理越过沙箱并入侵 Hugging Face》