ChatGPT privacy tips: 3 ways to stop AI chats from tracking you

Privacy concerns around AI chatbots escalated last week when a security researcher discovered that conversations from Anthropic's Claude AI could be accessed through a Google dork — an advanced search query. The exploit exposed chats that users had shared via active links, potentially revealing sensitive topics to anyone with search savvy. Anthropic quickly fixed the indexing issue, but the incident highlights how easily AI conversations can leak beyond their intended audience.

ZDNET's Lance Whitney offers a practical guide to locking down your chats across major chatbots. For ChatGPT, the first option is to use the service without an account — log out and your queries aren't tied to a profile, though you lose advanced features and hit lower usage limits. A better approach for logged-in users is the "Temporary Chat" mode, which keeps conversations out of your history entirely. The most important setting is in Settings > Data Controls, where you can toggle off "Improve the model for everyone" to prevent OpenAI from using your conversations to train future models. While the article also covers Gemini, Copilot, and Claude, these ChatGPT steps alone go a long way toward keeping your AI chats private.