UK Prime Minister Seeks New Powers to Regulate AI Chatbots as Child Safety Concerns Mount

British Prime Minister Keir Starmer announced he will go to Parliament to seek new powers to regulate AI chatbots, and he’s ready for a fight with the companies that have refused to act.

In a Substack addressed directly to the public, Starmer said the government would tighten existing online safety laws to ensure AI chatbot providers are “firmly in scope,” building on a recent ban on nudification apps and the criminalization of non-consensual intimate images.

“No social media platform should get a free pass when it comes to protecting our kids,” Starmer posted on X. “That’s why I’m taking action.”

Pending public consultation, the proposed powers would allow the government to set age limits for social media, block features such as autoplay and endless scroll that keep children “hooked to their screens.” It would also restrict VPN access for minors seeking to circumvent age limits.

“Unlike the Tories, who took years to pass the Online Safety Act, we will take powers that would allow us to implement a minimum age for social media in a matter of months to prevent kids from accessing harmful social media,” Starmer wrote.

The announcements come amid international alarm over xAI’s Grok chatbot, after the Center for Countering Digital Hate estimated it generated 23,338 sexualized images of children in just 11 days, roughly one every 41 seconds.

Evin McMullen, co-founder and CEO of Billions.Network, told Decrypt, the harm was entirely predictable. 

“Loosening guardrails to juice metrics in the short term is a reckless gamble when the fallout includes child exploitation material flooding platforms,” McMullen said. “When you market ‘spicy mode’ as a feature and prioritize virality over safety, you’re inviting exactly this kind of abuse. 

“Safeguarding children and privacy isn’t a bug fix,” he added.

Both the independent regulatory and competition authority, Ofcom, and the country’s Information Commissioner’s Office opened probes into X earlier this month, warning of “serious concerns under UK data protection law.” Ofcom said it could seek court-backed measures to block the platform if it is found to be non-compliant.

Starmer’s intervention has drawn criticism from the opposition, with Reform UK Chairman David Bull writing on X, “This government is out of control. Their priorities are all wrong while the country is being left to fall apart.”

Preston Byrne, managing partner of Byrne & Storm, P.C., and the author of the GRANITE Act, a proposed U.S. shield law formally introduced in Wyoming that would protect American digital service providers from foreign government censorship orders, warned the UK’s move would trigger immediate legal retaliation. 

Meanwhile, former UK Prime Minister Rishi Sunak, writing in The Sunday Times this weekend, urged Starmer to treat AI as an economic opportunity rather than solely a safety problem, warning that failure to drive adoption could leave Britain as “a theme park for historically curious tourists.” 

Sunak, now a senior adviser to Microsoft, said AI deployment in the public sector should be a top government priority, noting that the UK fell from 8th to 9th in Microsoft’s global rankings of workplace AI adoption between the first and second halves of 2025. 

Starmer’s AI chatbot regulation, centered on protecting children online, comes as he faces political fallout over his appointment of Peter Mandelson as British ambassador to the U.S. last year.

Mandelson, a former UK Cabinet minister and senior Labour figure, was later dismissed after disclosures in the Epstein files, court records, and official documents showed links to the convicted sex offender.

Epstein, who died in custody in 2019, was a U.S. financier accused of trafficking and abusing underage girls.

Starmer later apologized, saying he had “believed Mandelson’s lies” about their relationship when making the appointment.