WhatsApp owner Meta’s decision to add an AI assistant to the popular messaging app without an off switch has prompted an EU lawmaker to raise concerns with the European Commission (EC) over user control and digital safety.

In a post on X (formerly Twitter), MEP Veronika Cifrová Ostrihoňová said she had submitted a written question to the EC, the EU’s executive body responsible for enforcing laws and regulating tech, to clarify whether the AI complies with EU rules.

Ms Ostrihoňová, who belongs to the social-liberal Progressive Slovakia party, said she took action after several users contacted her with their concerns over WhatsApp’s recent update.

Meta began rolling out its AI assistant to WhatsApp users in Europe in March, following months of regulatory wrangling over data protection.

The tool, Meta AI, is available on WhatsApp for Android in the UK. An iOS release is expected this week.

Users can access Meta AI in several ways: through a new button in chats, by typing “@MetaAI” in a message, or via the search bar.

The chatbot works in both one-on-one and group chats. However, in Europe, it’s limited to text responses—unlike in the US, where it can also create images and stylised selfies.

While the UK is no longer part of the EU, it still operates under GDPR-style data rules. Last year, the UK’s Information Commissioner’s Office urged Meta to pause its AI training plans over how it was handling user data.

Much to the outrage of users and privacy campaigners, Meta later tweaked its opt-out process and went ahead with the UK launch, though the ICO said it would continue to monitor the situation.

Meta faced similar pushback from Ireland’s Data Protection Commission in the EU. This led the company to confirm that the version of Meta AI available across Europe has not been trained on local user data.

Even so, the rollout has prompted significant backlash, most recently in Ireland, where users were caught off guard by the tool’s sudden appearance. Judging by the deluge of negative posts on social media, WhatsApp users are scrambling for ways to turn off the feature, urging others to switch to rival apps including Signal, and voicing fears about their data being used to train the AI.

Meta defended the integration of the AI, insisting the feature is entirely optional.

“Meta AI is like any other feature,” Joshua Breckman, Communications Director, International, WhatsApp, told the Standard. Comparing it to existing tools such as Communities, Channels, disappearing messages and chat locks, he concluded that “some people will love it and use it, and some people won’t — and that’s absolutely fine”.

Mr Breckman emphasised that the company’s priority is preserving privacy. He said users’ messages remain end-to-end encrypted. “WhatsApp can’t read your personal messages,” he said, adding that it’s “really distinct” in the app when you’re interacting with the AI.

Despite recent complaints about the inability to disable the tool, Mr Breckman said most feedback has been positive. “The biggest question we get from people is how to turn it on,” he said.

Could users choose a version of WhatsApp without AI? Mr Breckman said: “There’s a non-AI version of WhatsApp if they don’t click on the button.”

He also described Meta AI as making powerful AI tools accessible to everyone.

“We’re giving people the option to interact with an AI that many won’t be able to afford or access elsewhere — and we’re putting it in a place they already know, trust and understand,” he said.