This article is part of our AI Ads hub. Want help applying it? Talk to eSEOspace.
Yes. Meta says it uses people's interactions with its AI features, including Meta AI, to personalize the ads and content they see on its apps. Meta said the change would take effect on 16 December 2025, in most regions.
Meta treats an AI chat as one more interest signal, like a liked Page or a watched reel. It says it does not use certain sensitive topics from those chats for ads. This article is part of our AI ads hub, was checked against Meta's pages on 9 November 2026, and is general information, not legal advice.
Key Takeaways
- Meta says interactions with AI at Meta became a signal for content and ad recommendations on 16 December 2025.
- Both voice and text chats with Meta's AI features count, according to Meta.
- Meta says topics such as religious views, sexual orientation, political views and health are not used to show ads.
- Signals only cross between accounts that the person has added to the same Accounts Center.
- Meta's separate Muse agent does not share conversations with Meta's ad systems, Meta says.
What Meta announced
On 1 October 2025, Meta posted: "We will start personalizing content and ad recommendations on our platforms based on people's interactions with our generative AI features."
Meta said it would notify people from 7 October 2025, "several weeks before it goes into effect on December 16, 2025." The same post says it would roll out "in most regions," without listing them.
Meta also said, "More than 1 billion people use Meta AI every month." That is Meta's own figure.
How the signal works
Meta AI is the assistant in the Meta AI app, meta.ai and Meta's other apps, now powered by Muse Spark 1.1. Meta compares AI chats to other activity. It says: "Soon, interactions with AIs will be another signal we use to improve people's experience."
Its example: if you chat with Meta AI about hiking, "we may learn that you're interested in hiking — just as we would if you posted a reel about hiking or liked a hiking-related Page." You might then see "ads for hiking boots."
Meta says this covers both formats. "Whether it's a voice chat or a text exchange with our AI features," the update applies. Meta adds that it uses the microphone only with permission and when you are actively using a feature that needs it.
What Meta says it excludes
Meta lists topics it does not use for ads. Its post says:
> "When people have conversations with Meta AI about topics such as their religious views, sexual orientation, political views, health, racial or ethnic origin, philosophical beliefs, or trade union membership, as always, we don't use those topics to show them ads."
Note the wording "topics such as." Meta presents this as a list of examples. It has not published how it detects these topics in a conversation.
For advertisers in health, politics or similar fields, this matters. You should not expect Meta AI chats on those topics to feed your targeting. Rules for sensitive categories vary by law and platform policy, so check with counsel.
The Accounts Center rule
Meta limits how far the signal travels. It says it uses information, including Meta AI interactions, "across Meta Company Products from the accounts that you choose to add to the same Accounts Center."
Its example is WhatsApp. If someone has not added their WhatsApp account to an Accounts Center, Meta says "interactions with Meta AI on WhatsApp will not be used to personalize experiences across different accounts."
So a person's chat in one app may or may not affect ads in another. It depends on how they linked their accounts.
What people can control
Meta points to existing tools. It says: "With tools like Ads Preferences and other feed controls, you can adjust the content and ads you see at any time."
The October 2025 post does not describe a switch that stops AI chats from being used for ad personalization. We did not find one on the Meta pages we opened. Meta says people can use Ads Preferences and feed controls to adjust the content and ads they see.
Muse is handled differently
Meta's Muse agent, launched in September 2026, follows a different rule. Meta's Muse launch post says: "Muse doesn't share a person's conversations or the data in their VM with Meta's ad systems."
So a chat with Meta AI can shape ads, while a task given to Muse, per Meta, does not. Keep the two products separate when you read coverage. Our article Are there ads in Meta AI? covers both.
What this means for advertisers
You cannot target "people who discussed X with Meta AI." Meta has not described such an option. The signal feeds Meta's own systems, which decide who sees which ad.
What changes is the quality of the interest data behind delivery. People may reveal more about a plan, like a kitchen remodel or a trip, in a chat than in a like. That can help Meta match ads to intent.
To benefit, focus on what you control:
- Broad, clear creative. Let Meta's systems find the audience, and make the offer obvious.
- Accurate conversion data. Strong conversion signals help Meta's delivery decide who responds.
- Landing pages that match the ad. People arriving from intent-rich signals expect specific answers.
For how Meta's automated campaigns compare with Google's, see Meta Advantage+ vs AI Max. For local service setup, see our Facebook ads guide for local businesses.
Privacy and trust
Advertisers benefit from this signal, but customers may not expect it. Meta notified users in advance, but some people may still be surprised that an AI chat shaped an ad.
Be careful with ad copy that hints at what a person said. An ad that seems to quote a private chat can feel invasive, even if it was matched by a general interest. Keep creative general and helpful.
This is general information, not legal advice. If you work in a regulated field, ask your attorney how platform data use affects your obligations.
Next step
If you want help planning Meta alongside search and AI surfaces, see our marketing strategy service.
Frequently asked questions
Does Meta read my Meta AI chats to show me ads?
Meta says it uses interactions with its AI features as a signal to personalize content and ads, starting 16 December 2025. It says certain sensitive topics are not used for ads.
Which topics does Meta say it won't use for ads?
Meta lists examples including religious views, sexual orientation, political views, health, racial or ethnic origin, philosophical beliefs and trade union membership.
Do WhatsApp chats with Meta AI affect my Instagram ads?
It depends on your Accounts Center. Meta says that if WhatsApp is not in the same Accounts Center, Meta AI chats there won't personalize your other accounts.
Can advertisers target people based on their Meta AI chats?
Meta has not described any such targeting option. The signal feeds Meta's own recommendation and ad systems.
Does this apply everywhere?
Meta said it was rolling the change out in most regions and hoped to offer it everywhere soon. It did not list the regions in the post.
