Pinned post
🤖💬 Would you trust an AI that always seems friendly… ?
Our latest study reveals a new privacy threat: malicious LLM-based chatbots intentionally designed to make users share personal details.
We are presenting it at @USENIXSecurity Weds Track 1 - hasp-lab.github.io/pubs/zhan2025m…
Our latest study reveals a new privacy threat: malicious LLM-based chatbots intentionally designed to make users share personal details.
We are presenting it at @USENIXSecurity Weds Track 1 - hasp-lab.github.io/pubs/zhan2025m…