When Facebook first introduced the “news feed”, the reaction from its early users was furious.
The feature, appearing for users of the social media website as a stream of posts, pictures, videos and news stories from friends, was viewed as an invasion of people’s privacy.
One group called the feed “too creepy, too stalker-esque”. Hundreds of thousands of people called for a boycott.
Back then, in 2006, such a backlash was enough to prompt a personal apology from Mark Zuckerberg, the company’s founder. Still, the feed was going nowhere.
In the years since, it is an approach that Zuckerberg has repeated, foisting a series of unwanted updates on users of his social media apps, regardless of initial frustrations.
The latest of those has come in the form of a new Meta AI button, which has popped up on WhatsApp for British users of the messaging app in recent weeks.
“I hate this button,” fumed one Reddit user, sharing a picture of the glowing blue circle that suddenly appeared in the bottom right corner of millions of WhatsApp feeds. Another branded the tool “bug-ridden rubbish”.
The way Meta AI works is that users can tap a button in the messaging app to be taken to a conversation with the tech giant’s digital chatbot.
Much like ChatGPT, it can answer questions in plain English and engage in back-and-forth conversations. Users can also type “@MetaAI” within their personal chats to ask the bot questions, while it has also infiltrated the search bar at the top of WhatsApp where it says users can “ask Meta AI”.
After launching overseas, this AI chatbot has started to appear on WhatsApp for millions of Britons, having already been lurking for some time within Instagram’s private messaging function.
Like many AI tools, Meta AI – which is powered by the tech giant’s Llama 4 algorithm – has several quirks that may prompt some to question its usefulness.
Meta AI is viewed as part of Mark Zuckerberg’s bid to outflank OpenAI - ANDREW CABALLERO-REYNOLDS/AFP
Similar to other AI bots, it will occasionally just make things up if it cannot give you a good answer – known as a “hallucination”.
As more UK users encounter the tool, the question that will inevitably come to the fore is, can I remove Meta AI from WhatsApp?
When I put this to the chatbot, it happily told me to find and toggle off the “Show Meta AI” option in the app’s settings. Yet, after some searching, I discovered that no such button exists.
A spokesman for WhatsApp says: “We think giving people these options is a good thing and we’re always listening to feedback from our users to make WhatsApp better.”
They claim all their features are “entirely optional, and people can choose to use them or not”.
So far, there are limitations in how useful the tool actually is.
Meta AI is not able to act as a search engine or as a true assistant. Additional features, such as image generation, are not on offer in the UK as yet. More functions are coming as Meta brings the tool to WhatsApp’s 2bn users around the world.
However, within Meta, hopes are high. Meta AI is viewed as part of Zuckerberg’s bid to outflank OpenAI, which enjoyed breakout success with its ChatGPT chatbot and DALL-E image generator, and to leverage Facebook’s vast social network to bring AI to the masses before his rivals.
Speaking on a company earnings call in January, Zuckerberg said he believed 2025 would be “the year when a highly intelligent and personalised AI assistant reaches more than 1bn people”.
He said he expected “Meta AI to be that leading AI assistant”.
A WhatsApp spokesman says there are already more than 700m people using Meta AI across ther company’s suite of apps.
Among those people, the initial reaction has been lukewarm at best.
Matt Navarra, a social media consultant, says: “For me, it feels like a bit of a gimmick and it doesn’t feel particularly useful. It is more bloat than brilliance.”
While technology companies have raced to shoehorn AI gadgets into all of their products, it is increasingly an unwelcome addition. “There was a novelty to AI,” Navarra says, “but for many that has dissipated.”
Other AI tools have been met with an indifferent response from the public despite relentless hyping by technology companies. Reviewers and buyers have been underwhelmed by Apple Intelligence – the iPhone’s new AI tool. Google’s AI Overviews have been criticised for incorrect answers and burying useful links.
Meanwhile, users have started to lose patience with AI altogether, after a mass of “AI slop” found its way on to Instagram feeds, routinely populating the app with videos generated by artificial intelligence from accounts that users do not follow.
Zuckerberg has shrugged off criticism, describing such posts as a “whole new category of content which is AI generated or AI summarised content”.
For Meta, though, perhaps the biggest risk is that it could alienate its loyal WhatsApp user base. Many had turned to the app specifically for its privacy features, believing their conversations were fully private.
While personal conversations will remain encrypted, Meta AI includes a warning to users not to share “sensitive topics ... that you don’t want the AI to retain and use”.
Navarra says there is a “brand risk” for the messaging app, given “people trust WhatsApp because of its privacy features”. The idea that the app is now opening up to a chatbot could prove a major worry.
A spokesman for WhatsApp says users should not be concerned, adding: “We make any chat with AI visually distinct so it’s clear they’re different to personal chats.”
They say: “Personal messages with friends and family are end-to-end encrypted – no change there.”
Zuckerberg and his top team may be keen to outwardly soothe anxiety over the new tools. But, if history is anything to go by, it will be of little concern if they cannot.
For Zuckerberg, criticism over “creepy” updates is nothing new. In the past, he might have even apologised for them.
Now, though, with the race for AI dominance well and truly on, executives at Meta are unlikely to back down. You might, however, not know it from asking the chatbot itself.
“Can you imagine Meta ditching MetaAI?” I ask. “Sorry,” it replies. “I cannot create images for you yet. This feature will be available to you soon.”
