Meta is expanding how it personalizes your experience across Facebook, Instagram, and WhatsApp — and now, that includes your conversations with its AI assistant.

Beginning December 16, Meta will start using topics from your text and voice chats with Meta AI to tailor the ads, suggested posts, Reels, and Facebook groups you see.

For example, if you chat with Meta AI about hiking, you might later see hiking-related Facebook groups, posts about local trails, or ads for hiking boots. Meta says this is similar to how it already uses your likes, follows, and other activity to infer your interests.

However, the company says there are limits. Sensitive topics — such as religion, sexual orientation, political or philosophical beliefs, health information, or union membership — will not be used for personalization. “We have existing policies around information that people might consider sensitive, and those will continue to apply,” said Christy Harris, Meta’s head of privacy, during a press briefing.

While you’ll still be able to adjust your ad preferences in settings, you won’t be able to opt out of Meta using your AI conversations for personalization.

If your Facebook, Instagram, and WhatsApp accounts are connected through Meta’s Accounts Center, the company may use AI chat data from one app to influence recommendations or ads in another. This includes one-on-one AI chats in WhatsApp and Messenger — though Harris emphasized that encrypted conversations remain unaffected by this update.

Meta will begin notifying users about the change on October 7, with a global rollout planned — excluding the UK, the EU, and South Korea, where regulatory approvals are still pending.

Source