ChatGPT Ads Target Your Phone Number, Email

Joseph Cox exposes how ChatGPT’s ad system lets advertisers target individual users by phone number, email, or even a device identifier and inject ads directly into chat sessions. Cox demonstrates that someone with only a few minutes of effort can upload contact data into OpenAI’s ad tools and serve tailored ads tied to a user’s prompts and identity; he tested this himself by creating a benign ad without using other people's personal data. The reporting warns that most users likely don’t realize an ad in their chat could be the result of a direct upload of their phone number or email into the platform’s targeting systems.

The piece also highlights new research showing ad networks are being used to deliver potent, zero-click mercenary malware, confirming that online advertising can be an attack vector as well as a targeting mechanism. The combination of identity-based ad targeting inside AI chat sessions and documented malware delivery through ad infrastructure raises both privacy and security risks for users whose contact data circulates in advertiser lists. The coverage notes that detailed targeting capabilities extend to prompts, phone identifiers, and standard contact fields, and that these practices are operable through OpenAI’s systems.