How to turn off Meta AI in WhatsApp? You can’t, but...
Noticed an AI assistant is now in your WhatsApp? Meta’s AI feature is now being rolled out in the EU and you can’t turn it off – but you can stop the AI invasion by opting for privacy-focused alternatives.
Update April 23, 2025: Last week Meta announced that it plans to start training its AI models using EU citizen’s public posts, comments, and any interactions with Meta’s chatbots. If you’re a Meta user in the EU you will receive email and in-app notifications to inform you that Meta will use your public data for AI training. This notification has a link to an objection form where you can object and opt out of having your public data used - if you’re still using Meta’s platforms we’d highly recommend opting out.
Follow these steps to object to Meta using your public data for AI: Under the heading What you can do click on Object this will take you to the objection form. From there enter your email address and click Submit.
How to turn off Meta AI
For users in the US, Meta AI has been around since 2023, and now it’s being rolled out globally. If you’re based in Europe you might be surprised to see Meta’s AI chatbot recently appeared in your WhatsApp. And no, you didn’t enable the feature by mistake – Meta has decided to turn it on for everyone in WhatsApp. This is the start of Meta’s AI roll-out, and soon its other apps - Instagram, Facebook, and Messenger - will also include AI features.
If seeing the blue circle button in the bottom right-hand corner of your WhatsApp is making you concerned for your message privacy, annoying you because it can’t be moved or turned off, or simply because you didn’t enable the feature, keep reading.
There’s currently no way to turn off AI in WhatsApp but there are other options you can consider which will allow you to message your friends and loved ones with better privacy, and no AI features.
When you open up the AI bot in WhatsApp, a message appears. While Meta says that, ‘Meta AI is an optional service from Meta that uses AI models to provide responses’, it’s worth highlighting that it’s only optional to use and engage with - it wasn’t an optional feature to turn on, and it’s not optional to remove. Screenshots: WhatsApp
Meta AI is here to stay, and as it mentioned in its official announcement of the roll-out in Europe, parent-company Meta plans to expand its offering over time. For now, Meta will not let you turn off Meta AI in WhatsApp. So if you choose to keep using Meta’s instant messenger, you will have to get used to the annoying blue AI circle.
Ditch Meta and choose privacy-focused alternatives
If you don’t want AI features in your instant messenger or social platforms, you will need to stop using Meta products, and opt for other alternatives. You might be worried about leaving Big Tech services like WhatsApp behind, but you will be happy to hear that globally people are now choosing to take back their privacy and stop using services offered by tech giants.
As people are becoming aware of how Big Tech like Meta and Google use their data, they are purposefully ditching these Big Tech-owned products entirely by “deGoogling”, moving away from Meta and Microsoft, and choosing European replacements as these often offer better privacy protection, especially in tech.
Best WhatsApp Alternatives
-
Signal
Signal is the WhatsApp alternative best for privacy. It’s now the recommended messaging app to use by the US Agency CISA and the Swedish Armed Forces. Signal offers excellent privacy and security due to its quantum-resistant encryption, open-source code, and minimal data collection. Signal is one of the more popular privacy-focused messengers as it’s free to use, comes with great features, and available for download on iOS, Android, and desktops. Signal does require a mobile number for sign-up.
-
Threema
Threema is a Swiss-based instant messaging app that is fully anonymous as you can sign up without a phone number or email address. Built for privacy, Threema ensures your messages are sent end-to-end encrypted, it’s also open source, doesn’t collect your data, and is GDPR compliant. Threema costs a once-off fee of 4.99 Euros for private users and it’s available for download on iOS and Android. It can also has a desktop app available.
-
Element X
Element X is another end-to-end encrypted messaging app that’s open source, with a strong focus on privacy and security while also offering a user-friendly interface with all the necessary features. Element X is available for mobile download on iOS and Android. Element also offers a desktop app.
Want to know more? Check out our in-depth guide of the best WhatsApp Alternatives.
Facebook and Instagram Alternatives
Soon, Meta AI will also have features on Facebook and Instagram. But even before the introduction of AI into these platforms, these social sites have never been the best for privacy. Meta offers its products for free because it profits off your data. This means that by using the tech giant’s social platforms, you have become the product. So it’s no surprise that FaceBook knows and collects so much about you! And now with the introduction of AI and needing to train Large Language Models (LLMs) off user data, your data is even more valuable. Below, we look at two alternatives to Meta-owned social platforms that are free to use, growing in users, and are not focused on collecting your data to increase profits. While these platforms do not use your data for AI training, unfortunately, there is no guarantee that your data isn’t scraped by others for AI.
-
Mastodon
Mastodon’s header text on their website, “Social networking that’s not for sale” summarizes what the social platform is all about. Mastodon is a German-based, non-profit social platform that’s decentralized, open source, and free to use. It does not serve you ads, or use sneaky methods to push specific profiles or news into your feed. Because it’s run as a non-profit, it doesn’t need to drive profits (by collecting and selling user data) to keep shareholders happy. Instead, there’s more reassurance that the main goal of the platform is to benefit the users.
-
Bluesky
Bluesky is a social platform similar to X, that’s rising in popularity, and growing in users as people are starting to flee X. Bluesky describes itself as a “social media as it should be”. Like Mastodon, this social platform is also decentralized and open source. In terms of use and features, it’s similar to X and free to use.
When did Meta AI start
Meta first launched its AI features to users in the US in 2023 and since then, it has been pushing to roll it out globally. But introducing AI in Europe has been a bit more tricky for the tech giant. Last year, Meta started informing users within the EU/EEA of upcoming changes to its privacy policy which if applied would allow it to start using public user data from comments, status updates, photos, and interactions from Facebook and Instagram for AI training.
Meta’s plans to introduce AI got put on hold by NOYB as it raised questions of whether it was in compliance with the EU’s GDPR. One of the big issues raised was that users would not be given the choice to opt in or opt out – but rather got forcibly opted in by default.
Because of these concerns, Meta faced backlash notably from the privacy NGO, None of Your Business (NYOB) who filed complaints against the tech giant in 11 European countries. The Irish Data Protection Commission also advised Meta to delay its plans to use user data for AI training, and with that, the introduction of Meta AI was put on hold.
But now, almost one year later, Meta has started rolling out the AI writing assistant. What’s surprising to see is that this AI assistant is now being rolled out across its platforms in 41 European countries, and you can’t turn it off. More concerningly, users in the EU haven’t been given an option to “opt in” or “opt out”.
No option to disable AI
It’s worrying to see that the introduction of the AI tool in WhatsApp (and other platforms to follow) is not the choice of its users and can’t be disabled. While Meta AI was not trained using EU users’ data, and Meta has said that personal chats are not used to train its AI Models, the interactions you have with the bot could be used to improve future versions of the AI model.
In addition, even if a Big Tech says, they are not using your data, you can never be one hundred percent certain. At least with Google’s Gemini, there was the case that Gemini itself said it would use personal Gmail emails for composing replies.
The abrupt appearance of the AI tool in WhatsApp is especially concerning for those who value their privacy and for those who are aware of the Big Tech’s greedy data collection practices. Additionally, knowing that AI models are improved and trained off of user data, it’s not reassuring to have this AI tool integrated without choice.
Around Europe, many are sharing their concerns and frustration of finding the AI bot in their private messaging app, and not being able to turn it off, for instance onBluesky and Reddit.
WhatsApp users in Europe have shared their concerns over Meta AI’s arrival. From complaining over not having the choice to opt out, to the interface having the circle AI button permanently on the screen, it is evident many Meta users are not welcoming Meta AI. Screenshots: Bluesky and Reddit.
This is just the start
The AI race is on, and now tech giants are quietly updating their terms of use and privacy policies to opt users into having AI features integrated into their services, and for user data to be used for AI purposes by default. They do not give you the chance to willingly enable the AI feature and they do not explicitly ask for your permission to collect and use your data. And we are seeing this more and more. Whether it be opting you into having an AI assistant in your Gmail inbox, LinkedIn opting you into having your data used for AI training, or almost all social media platforms using your data for AI. It’s widely known that they use our data for advertising purposes, but now, with the development of AI, human data has become even more valuable.
What’s Next
Using privacy-focused platforms that don’t abuse your data and introduce dodgy AI features is the only way you can reduce AI’s extensive reach into your life.
Moving apps and services can feel overwhelming, especially when we are inclined to go for the most popular services, for example, WhatsApp, Gmail, and Instagram. But ditching these data-hungry services that are profiting off of you will give you peace of mind knowing your private communications and data are protected.
It’s important to remember that moving providers doesn’t happen overnight, so don’t try to change everything at once. It’s a slow and steady process that takes effort but is well worth it. (I promise!)
Want more privacy?
WhatsApp is not the only communication tool now riddled with AI. If you use popular email providers like Gmail, AI is even in your mailbox. The best and easiest solution to protect your mailbox is to switch to Tuta Mail.
Tuta Mail is an end-to-end encrypted email provider based in Germany that is designed to give you the best privacy and security for your communications.
Tuta Mail is free to use and does not serve you ads or collect your data. Switching to the secure email provider is simple, and it is available for fee on Android, iOS, Linux, Windows, MacOS, or via your web browser.
