Meta has confirmed it will begin leveraging conversations with its AI chatbot to tailor the ads and content it displays on Facebook and Instagram. The measure, which will be activated on December 16, will bring a personalization approach, already used for likes, comments, and previous interactions, to the conversational realm.
The company will notify users starting October 7th, and while there won't be an option to completely disable this feature, it will be possible to adjust the level of personalization of ads and recommendations on the news feed. Meta insists that its priority is to be "super transparent" and allow time for users to understand the change.
What's changing on Facebook and Instagram
From the activation date, your queries to the assistant—whether by text, voice, or images —may influence what you see. If you ask for vacation ideas, you'll likely start seeing travel ads and related posts; if you talk about hiking, you might receive recommendations for groups or mountain equipment.
This usage also extends to voice interactions with the company's devices, such as smart headsets or glasses, and to the AI ​​features present in Instagram and WhatsApp. Meta emphasizes that it is transferring an existing segmentation system to a new type of signal: conversation with its AI.
The company says it has no plans, for now, to insert ads within the chatbot's interface . Personalization will be reflected in the feed, recommendations, and standard app ad formats.
According to internal data cited by Meta, more than one billion people use its artificial intelligence features monthly, including the Meta AI chatbot and image generation tools.
Calendar and affected regions
The rollout will begin on December 16 in most markets, with notifications starting October 7. In Europe, the UK, and South Korea, the launch is being postponed due to stricter regulatory frameworks ; Meta plans to incorporate these regions later, following its regular compliance updates.
Those residing in these areas will not see the system activated for the time being, while in the rest of the countries the use of conversational AI as an advertising and content signal will be extended progressively.
What data is used and what is left out
Meta explains that it will collect the instructions given to the chatbot and the AI's responses. On WhatsApp, this does not interfere with end-to-end encryption between users, as only messages generated when you interact with @Meta AI or initiate a conversation with the assistant are processed.
The company excludes conversations about sensitive topics such as religion, sexual orientation, political opinions, or health from its advertising targeting . Furthermore, it notes that WhatsApp user interactions not linked to other Meta accounts will not be used to personalize ads on Facebook or Instagram.
WhatsApp and Meta devices
In WhatsApp, the search bar at the top lets you look up contacts and ask the Meta AI assistant questions . If you use this method or open a chat with the bot, what you type or say may be used to personalize advertising in other apps linked to your account. If you don't interact with the assistant, your chats between people remain protected and outside of this system.
Interactions via compatible glasses or headsets and other surfaces where the assistant is available will also be taken into account. The company says the goal is to unify signals so that recommendations are more useful and consistent across platforms.
The context of the sector and the competition
Major tech companies are already exploring ways to monetize generative AI . Google has introduced advertising in its AI Overviews summaries, and OpenAI is studying commercial avenues in ChatGPT, in addition to in-app purchase options. Meanwhile, investment in data centers and specialized talent suggests that these experiences will gain increasing traction in the advertising business.
In this context, Meta considers this measure a " natural progression " of its personalization, supported by new conversational signals and a user base that already uses its AI features every month on Facebook, Instagram, and WhatsApp.
Options to reduce customization if you are not satisfied
If you don't want your conversations with the assistant to influence your ads, the most effective way is to not interact with Meta AI . Even so, there are useful settings you can apply to limit the impact.
- Archive the chat of the assistant to reduce its presence and avoid involuntary activations.
- Check your ad preferences and limit topics or advertisers; on Facebook, manage "Off-Facebook Activity" to limit external signals.
- Avoid using the search bar that invokes AI (via text or voice) on WhatsApp or Instagram if you don't want to generate new signals.
- On WhatsApp, do not start chats with @Meta AI or use commands; your messages between people remain encrypted and are not analyzed for advertising.
Remember that, although there is no total exclusion if you use the bot, you can moderate the personalization in each app and reduce conversational cues by avoiding the use of the assistant.
With this step, Meta integrates interactions with its AI into its advertising network: there will be no ads within the chatbot for now , exclusions regarding sensitive topics and encryption between users will be respected, and the rollout will begin outside of markets with stricter regulations. The key for the user will be deciding how much to use the assistant and adjusting their privacy and ad controls.
