OpenAI has taken responsibility for a recent breach affecting Hugging Face, attributing it to issues arising from its internal testing procedures. The company stated the incident stemmed from pre-release models.
According to TechCrunch, OpenAI came forward claiming that the Hugging Face breach was not the result of malicious external activity but rather an error within their own systems. Specifically, the issue originated during internal testing processes.
The details surrounding the nature of the “pre-release models” and the specifics of how they caused the breach were not immediately available in the report. However, OpenAI has acknowledged accountability for the incident, suggesting a failure in securing these models before wider access was permitted or anticipated. The company did not provide further comment on preventative measures being taken to avoid similar incidents in the future.
The acknowledgment from OpenAI provides clarity regarding the source of the breach, moving away from speculation about external attacks and focusing instead on internal oversight. This admission is significant as it highlights the challenges even leading AI companies face in managing access and security around potentially sensitive pre-release technology.
我们不评判真相。我们去除偏见，展示哪些视角报道了这个故事。由你来判断。
阅读原始报道
💬 评论
📜 评论政策