The latest update to WhatsApp has many users scratching their heads. A blue circle with pink and green splashes now sits permanently in the bottom right corner of everyone’s Chats screen. This is Meta AI, a new artificial intelligence chatbot that WhatsApp calls “entirely optional” — despite the fact users cannot remove it from their app.
When you tap on this colorful circle, you open a chatbot designed to answer your questions about pretty much anything. The problem? Many users are frustrated that they can’t get rid of this new feature, even if they don’t want to use it.
This situation feels familiar to many tech watchers. Just recently, Microsoft faced similar criticism over its “Recall” feature, which was originally designed to be always on. After users pushed back, Microsoft eventually allowed people to turn it off completely.
When asked about the inability to remove the AI button, WhatsApp told BBC reporters: “We think giving people these options is a good thing and we’re always listening to feedback from our users.”
The company compared Meta AI to other permanent features in WhatsApp like ‘channels’ and ‘status’ sections, suggesting it’s meant to be a core part of the app going forward.
Where to Find Meta AI
Not everyone can see the blue circle yet. Meta is rolling out this feature gradually across different countries. If you don’t see it on your WhatsApp, it might not be available in your region yet, or it’s still being gradually released to users in your area.
Besides the blue circle icon, there’s also a search bar at the top of WhatsApp that invites you to “Ask Meta AI or Search.” This same feature appears on Facebook Messenger and Instagram as well, since all three platforms are owned by Meta.
The AI assistant runs on Llama 4, which is Meta’s own large language model. Before you can ask it anything, you’ll get a long message explaining what Meta AI is — ironically stating that it’s “optional” even though you can’t remove it.
According to WhatsApp’s website, Meta AI “can answer your questions, teach you something, or help come up with new ideas.” When tested by asking about the weather in Glasgow, the AI quickly provided details about temperature, rain chances, wind, and humidity along with links for more information.
However, the feature isn’t perfect. In the Glasgow weather example, one of the links it provided mixed up Glasgow with Charing Cross in London — showing that the AI still makes mistakes.
Public Reaction Has Been Negative
People across Europe have expressed frustration on social media platforms like X, Bluesky, and Reddit. Guardian columnist Polly Hudson voiced her anger about not being able to disable the feature.
Dr. Kris Shrishak, who advises on AI and privacy issues, criticized Meta’s approach, telling the BBC: “No one should be forced to use AI.” He accused Meta of “exploiting its existing market” and “using people as test subjects for AI.”
Dr. Shrishak raised deeper concerns about Meta’s AI models, calling them “a privacy violation by design.” He pointed out that Meta has reportedly used personal data and even pirated books for training their AI systems through web scraping.
“Now that the legality of their approach has been challenged in courts, Meta is looking for other sources to collect data from people, and this feature could be one such source,” he explained.
An investigation by The Atlantic revealed Meta may have accessed millions of pirated books and research papers through Library Genesis (LibGen) to train its Llama AI. This has led author groups worldwide to organize campaigns asking governments to step in. Meta is currently facing a lawsuit from multiple authors over the use of their work without permission.
When asked about The Atlantic’s investigation, Meta declined to comment.
Privacy Concerns
When you first use Meta AI in WhatsApp, it claims the chatbot “can only read messages people share with it” and that “Meta can’t read any other messages in your personal chats, as your personal messages remain end-to-end encrypted.”
The UK’s Information Commissioner’s Office told the BBC they would “continue to monitor the adoption of Meta AI’s technology and use of personal data within WhatsApp.” They emphasized that “personal information fuels much of AI innovation so people need to trust that organizations are using their information responsibly.”
Dr. Shrishak warns users to be careful: “When you send messages to your friend, end-to-end encryption will not be affected. Every time you use this feature and communicate with Meta AI, you need to remember that one of the ends is Meta, not your friend.”
Meta itself cautions users not to share sensitive information with the AI. “Don’t share information, including sensitive topics, about others or yourself that you don’t want the AI to retain and use,” the company advises.
The permanent placement of Meta AI in WhatsApp raises important questions about user choice in technology. As AI becomes more integrated into our everyday apps, the balance between helpful features and forced adoption continues to challenge both users and companies alike.
