- Meta starts using conversations with its Meta AI assistant on Facebook and Instagram to customize ads and content
- There is no opt-out
- Conversations in WhatsApp are not used for ads for now
Talking to Meta AI -Chatbot on Facebook or Instagram begins to affect the kind of ads and other content you see in a few months. The technology giant announced that the plan to use your interactions with the AI assistant to adjust what you see begins on December 16th. And you won’t be able to opt out of it. Meta begins notifying users of the update this month.
“Whether it’s a voice chat or a text exchange with our AI features, this update will help us improve the recommendations we deliver to people across our platforms, so they are more likely to see content they are actually interested in – and less of the content they are not,” Meta explained in its message. “For example, if you chat with META AI about hiking, we can learn that you are interested in hiking-offs we would if you sent a roll of hiking or liked a hiking-related page. As a result, you may be able to start seeing recommendations for water groups, posts from friends about trails or ads to hiking stands.”
Meta clearly wants people to think about its use of AI data as the same as its use of information you post in public. But behind the usual pitch of relevance lies a sharper edge: This is about using AI chats as another rich vein of behavior data, putting assistant conversations on par with likes, shares and clicking in the design of your digital experience and by making money. And the only way not to have Meta AI uses your data is to avoid interacting with the chatbot.
Or you can stick to WhatsApp, which remains excluded from Meta’s Administration Plan for now. But if you’ve ever asked Meta AI to suggest a restaurant or recommend a training routine, this exchange is to become part of the invisible scaffolding behind your feed.
The company says it will respect boundaries around sensitive categories such as politics, religion, sexuality and health, but the definitions of sensitive are hardly universal, and mislassification seems inevitable.
On the other hand, it is likely that only a few users are still under the illusion that their online experience is purely organic. But treating AI -Chat the same as by commitment or ad click is new. A chatbot is not a post. And when AI assistants become more natural and conversation, it blurs the mental line between talking to a tool and feeding a data collector. The illusion of intimacy can actually increase people’s willingness to share, even when the assistant calmly takes notes to the ad engine.
Meta is a pioneer in this project, but Google has talked about and experimented with AI ads in Gemini and from its AI listings, and Amazon uses conversations with his Rufus AI -Chatbot for similar purposes.
However, there is a geographical exception to Meta AI’s new role. The development is skipping Europe and the UK for now thanks to the data protection law as the EU’s GDPR. These rules slow down changes in the use of personal data. Meta works to make it happen there, but it can take some time.
Whether this is a good feature of meta can depend on whether people agree that asking a chatbot about advice on your kitchen remodel should mean to see related content later that day, or that it is the same as asking for a recipe. Meta is betting that most people will accept the convenience and will not object to the mechanics behind the scenes. But if you do not want Meta Ai to determine the kind of ads you see, you will ask elsewhere for ideas about your next purchase.
