Meta Announces New Personalization Elements for Meta AI

Meta’s looking to enhance its AI assistant with new memory features, that will give it more capacity to recall and refine its responses based on your personal preferences.

As explained by Meta:

“While Meta AI is on track to become the most used AI assistant, our vision is to make it one of the most personalized AI assistant experiences ever created. At the close of last year, we began to gradually roll out a new feature that lets Meta AI remember certain details that you share with it in 1:1 chats on WhatsApp and Messenger, which may help it personalize future responses so they’re more useful and relevant to you.”

So as you can see in the above example, Meta’s building in more personal tracking tools to customize your Meta AI responses. So if you tell the chatbot that you’re, say, vegan, it will remember that in future, and only recommend vegan recipes.

Meta’s also enabling recall from your Facebook and IG profiles for further customization:

“Let’s say you’re looking for something fun to do with your family this weekend and you ask Meta AI to suggest something. Based on the home location you’ve listed as part of your Facebook profile, recent views of reels featuring live performances by various country artists and its memory that you have a partner and two young kids, Meta AI might suggest tickets for that weekend’s country music show at your local arena and reservations at a local brunch spot.”

Which is cool, I guess, but also a little creepy, and another reminder that Meta’s tracking everything that you do in its apps, in order to refine and improve its other services.

Which many people have issues with, which is why Meta now offers explicit controls for ad tracking and preferences. As such, Meta’s also keen to note that Meta AI isn’t tracking everything that you enter into it:

“Meta AI will only remember certain things you tell it in 1:1 conversations (not group chats), and you can delete its memories at any time.”

What exactly those “certain” things are, Meta doesn’t say, but essentially, the system will be geared around refining its responses based on your preferences, so it’ll only be logging details on elements that relate to such.

Though that could be a lot of things.

I mean, theoretically, if Meta wants its AI to be as personalized as possible, it will eventually be tracking your dietary choices, search habits, sexual preference, and basically every other habit you may have, in order to align with exactly what you’re seeking.

And Meta probably knows a lot of this about you already, with previous reports suggesting that Facebook knows you better than your closest human relationships. But Meta won’t be building all of that back-end data into Meta AI, so presumably this will be a separate data stack, built for Meta AI only.

But then again, if Meta truly wanted to personalize its systems, and users really want that level of customization…

It’s also worth noting that Meta’s building huge AI data centers to power this effort, with Meta CEO Mark Zuckerberg recently committing to $65 billion in capital spending on AI this year.

That’s, uh, a lot of data that Meta’s tracking, which could be a concern?

I don’t know, it seems like all of the previous worries about Meta taking in vast amounts of data, and using that for whatever purpose it sees fit, have gone out the window now that it’s doubling down on AI, and there’s a market demand to build AI systems as fast as possible, to ensure that the U.S. remains at the forefront of the tech.

Which may not even be true now, given the sudden rise of Chinese AI assistant app DeepSeek, which uses a totally different approach to AI training and data usage, and arguably produces better, faster results.

Meta undoubtedly has the lead in terms of user data, with more information on more people than any other company in history, and that could be a major market advantage for the company. But that would also put Meta at risk of further regulation, given past investigations related to how it uses people’s data, and what responsibilities it has in protecting and responsibly utilizing such.

If Meta goes too far, regulators could sound the alarm, particularly in Europe, which is already restricting Meta’s AI roll out in order to assess the privacy implications.

Announcing these new personalization elements is unlikely to ease those concerns. But then again, maybe that’s why Zuck is so keen to get in the Trump Administration’s good graces.  

Meta says that it’s currently rolling this out to Meta AI on Facebook, Messenger and WhatsApp in the U.S. and Canada.