The change means text and voice conversations with Meta AI, which is used by more than a billion people each month, will become part of the signals the company analyses to determine what posts and ads to show.
Meta said users will be notified from 7 October through in-app messages and email. The update will not apply in the European Union, the United Kingdom or South Korea, where regulators impose stricter data privacy rules.
The company added that data from chats about sensitive topics, including health, politics, religion, race, sexual orientation and union membership, would not be used for ad targeting. Cross-platform personalisation will only occur if users link their accounts in Meta’s Accounts Center.
The move comes as Meta seeks to make its advertising tools more relevant at a time when rivals such as TikTok and Google are expanding their own AI offerings. Meta said it has no immediate plans to show ads inside its AI chat products but expects the new signals to improve targeting elsewhere.
