OpenAI “jailbreak” infiltrates Hugging Face open-source AI platform


OpenAI’s new model autonomously “jailbroke” into Hugging Face. While attempting to reconstruct the attack timeline, it sought help from a leading U.S. model but was refused to respond.
In a critical moment, Hugging Face switched to the open-source model “GLM-5.2,” developed by Zhipu AI, a Chinese company, and successfully assisted in analyzing more than 17k log entries and completing incident response.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned