Back
Mistral AIMistral API
In le Chat, we added a mitigation against an obfuscated prompt method that could lead to data exfiltration,…
AI summary
Written by AI from the official notes. Check them for exact details.Mistral AI enhanced le Chat to prevent a potential data exfiltration method using obfuscated prompts.
- Mitigation added against obfuscated prompt method
- No users were impacted by this issue
- No data was exfiltrated during the incident
Why it matters: Users of le Chat should be aware of this update for improved security against potential data risks.
Full release notes
- 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.