AI chatbots are leaking personal information, including phone numbers, to third parties without user consent. Researchers discovered that ChatGPT, Google's Gemini, and Anthropic's Claude retain and share user data in ways many people don't expect.
The problem stems from how these chatbots store conversation history. When you ask questions or share details, the bots record your inputs. These companies then use this data to train their systems and improve their models. Your phone number, email address, or other contact details can end up in their databases.
Third-party integrations make the situation worse. Many chatbot platforms connect to external services like scheduling apps, email providers, and payment processors. Your information flows between these systems automatically. You rarely see explicit consent requests before this data sharing happens.
What protects you varies by service. ChatGPT lets you disable chat history in settings, which prevents OpenAI from using your conversations for training. Gemini offers similar privacy controls through your Google account settings. Claude provides options to opt out of data retention. However, most users never find or adjust these settings.
The financial risk is real. Criminals who obtain your phone number can use it for social engineering attacks, SIM swaps, or SMS phishing. They can impersonate you to financial institutions or attempt account takeovers. Your number combined with other leaked data from breaches becomes more dangerous.
Steps to protect yourself: Review privacy settings in each chatbot you use. Disable chat history if you want to limit data collection. Avoid sharing sensitive information like bank account numbers, Social Security numbers, or full addresses. Use separate email addresses for different services when possible. Enable two-factor authentication on accounts linked to your phone number.
Consider whether you need to use these free chatbots at all. Paid tiers sometimes offer better privacy controls. Enterprise versions of ChatGPT and Claude don