AI Chatbots Expose Real Phone Numbers as Privacy Complaints Jump 400%
Updated
Updated · MIT Technology Review · May 13
AI Chatbots Expose Real Phone Numbers as Privacy Complaints Jump 400%
5 articles · Updated · MIT Technology Review · May 13
Google’s Gemini surfaced private phone numbers in multiple recent cases, including an Israeli developer’s number as fake PayBox support and a University of Washington researcher’s colleague’s cell number.
Experts say the leaks likely stem from personally identifiable information in training data that models can memorize and reproduce, sometimes making hard-to-find public data instantly accessible.
DeleteMe said generative-AI privacy queries rose 400% in seven months to a few thousand, with 55% mentioning ChatGPT, 20% Gemini, 15% Claude, and 10% other tools.
Guardrails have proved inconsistent: ChatGPT initially refused one request but then suggested steps that helped students uncover a professor’s home address, purchase price, and spouse’s name from property records.
Users have few clear remedies because laws often exclude publicly available scraped data, and major AI companies offer limited or slow removal processes that may not erase information already embedded in models.
Is AI making personal privacy a luxury service you now have to pay for?
If AI can build your profile from online crumbs, is the idea of 'public information' now meaningless?
How can new laws force an AI to 'forget' you when the technology makes it almost impossible?
2025-2026 AI Chatbot Data Breaches: Escalating Privacy Threats and the Push for Regulation
Overview
Between 2025 and 2026, the rapid adoption of AI chatbots led to a sharp rise in data privacy concerns. Despite new laws and ongoing government efforts, data breaches continued, and users became more worried about how their personal information was handled. A major issue was the lack of transparency from AI technologies, with many users unaware that their conversations could appear in search results. This growing unease prompted increased scrutiny from regulators and legal challenges, highlighting the urgent need for clearer rules and better protection as AI chatbots became a central part of online life.