Tracking Conversations: Measuring Content and Identity Exposure on AI Chatbots
Muhammad Jazlan, Ethan Wang, Yash Vekaria, Zubair Shafiq · 2026
Read Tracking Conversations: Measuring Content and Identity Exposure on AI Chatbots by Muhammad Jazlan, Ethan Wang, Yash Vekaria, Zubair Shafiq (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-27438.txt.
Description
Jazlan and three colleagues systematically measure web tracking on 20 AI chatbots, using a sensitive prompt in ordinary chats and, where offered, private chats. They check content exposure, including prompts, titles, links, and identifiers, and identity exposure, including names, emails, identifiers, cookies, and network fields. They find 17 of the 20 chatbots pass information to at least one outside party. Three share plain conversation text with Microsoft Clarity through replay. Fifteen share links or identifiers with advertising, analytics, or social endpoints, and several expose identity through widgets and replay tags, including hashed emails.
Catalog record
| CREATOR | Muhammad Jazlan, Ethan Wang, Yash Vekaria, Zubair Shafiq |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35199 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.27438