August 25, 2026, 6:54 am | Read time: 3 minutes
Regular users of ChatGPT leave countless digital traces. A free online tool now shows what personal information can be derived from these traces and how detailed the profile that AI systems create of their users can be.
A Personality Profile Emerges from Many Chats
Artificial intelligence has long been established in everyday life. Whether for work, travel planning, or personal questions, millions of people regularly turn to ChatGPT, Gemini, Copilot, and other systems. In doing so, they often share information about their interests, habits, or life circumstances.
This is exactly what the new online tool “AI Paper Trail” from the Swiss company Proton examines, reports the IT portal BASICThinking. The service aims to reveal what information users have disclosed during their AI usage. According to Proton, even inconspicuous details can be enough for AI providers to infer one’s profession, hobbies, relationships, or personal preferences.
Also interesting: Advertising on ChatGPT? There is a way out!
New Proton AI Focuses on Data Privacy Over Data Training
You can now chat with “Kendall Jenner” and “Paris Hilton” on WhatsApp
What the Tool Discovers About Users
Currently, AI Paper Trail supports the platforms ChatGPT and Claude. Users can export their stored conversation data and then upload it to the tool. The service then creates a detailed report on their digital footprint.
The analysis covers several areas. For instance, the system evaluates how easily a person can be profiled, what information about identity and life circumstances emerges from the chats, and what value this data might have from a provider’s perspective. Particularly interesting is that many insights do not arise from individual queries. Only the sum of numerous conversations over weeks or months provides a comprehensive picture of the personality.
Why the Self-Test Provokes Thought
The evaluation makes it clear how much information users often unconsciously disclose. Even those who never write account details, addresses, or real names in a chat can leave a meaningful profile through recurring themes and questions.
Proton therefore recommends using personal information in AI chats as sparingly as possible and regularly checking the privacy settings of the respective services. Where available, users should also disable options that use conversations for training AI models. Temporary or private chat modes can also help store less data permanently.
However, Proton does not act entirely without self-interest. The company itself offers an AI assistant and positions it as a privacy-friendly alternative to large U.S. providers. Nevertheless, AI Paper Trail provides an insightful look at how much modern AI systems can learn about their users from everyday conversations alone.