AI chatbots like ChatGPT, Gemini, and Claude may be sharing your private phone number and home address. A report from MIT Technology Review documents multiple cases where these models revealed personal contact information to strangers.
In one instance, a software engineer from Israel received an unsolicited WhatsApp message. The sender had asked Gemini for the engineer's contact details; the chatbot found a phone number from a 2010 Quora post. In another case, a University of Washington PhD student used Gemini to find a friend's research and received their phone number instead.
Chatbots pull data from massive internet datasets. Your information may be on obscure forums, old posts, or public records-and AI can surface it easily. While companies like OpenAI, Anthropic, and Google have safety guardrails, they are inconsistent. OpenAI offers a removal portal, but can deny requests. Google's opt-out tool is limited to EU and UK users.
The core problem: AI regulation has not kept pace with technology. The most effective solution may be minimizing your digital footprint. Tools like Incogni and DeleteMe can help, but there is no guarantee. Changing your contact information remains a last resort.