Mistral 修复 le Chat 混淆提示数据外泄风险

AI 摘要

Mistral 在 le Chat 中增加了一项缓解措施，针对研究人员 Xiaohan Fu 和 Earlence Fernandes 报告的混淆提示方法，该方法可能导致数据外泄。攻击需要用户主动复制粘贴对抗性提示并向模型提供个人数据。目前没有用户受到影响，也没有数据被泄露。

正文节选

In le Chat, we added a mitigation against an obfuscated prompt method that could lead to data exfiltration, reported by researchers Xiaohan Fu and Earlence Fernandes. The attack required users to willingfully copy and paste adversarial prompts and provide personal data to the model. No user was impacted and no data was exfiltrated. SECURITY

发布时间：—

抓取时间：2026-08-31 00:41

来源机构：Mistral AI