Artificial intelligence chatbots could soon face the same legal restriction as social media platforms under plans outlined by the UK government to tighten online safety laws, as Prime Minister Keir Starmer warned technology firms they will be held accountable for illegal AI-generated content.

Under the proposals, providers of AI chatbots would be brought within the scope of the Online Safety Act 2023, meaning they must comply with the same obligations as social media companies to prevent illegal content or face enforcement action.

Ministers will launch a public consultation on whether to introduce a minimum age for social media access and restrict potentially addictive features such as infinite scrolling. Spain, Greece and Slovenia have explored similar restrictions after Australia became the first country to block under-16s from social media.

The government also intends to close what it describes as a legal loophole in the Online Safety Act that currently excludes one-to-one chatbot interactions from safety rules. The technology minister, Liz Kendall, said the gap must be addressed, warning that some children are forming relationships with AI systems that were not designed with child safety in mind.

Tighter restrictions 

Under the plans, tech firms would be required to ensure their systems comply with British law. Ministers are also consulting on measures such as automatic data-preservation orders when a child dies, limits on “stranger pairing” in gaming and restrictions on sending or receiving nude images.

Starmer said the law must keep pace with rapidly evolving technology and that Britain should lead on online safety.  Officials said they would act swiftly to make the online world safer as part of broader reforms that could include age limits, VPN usage curbs and data preservation measures for children’s accounts.

“We will not allow tech companies to wriggle out of responsibility for vile illegal content created by AI. The law must keep pace with the technology, and Britain should lead the world in keeping children safe online,” Starmer said.

Critics have welcomed the focus on child protection but warned enforcement could be challenging and that restrictions might push harmful behaviour to less regulated corners of the internet.

• Spain to probe X, Meta and TikTok over AI-generated child sexual abuse content
• EU launches inquiry into Elon Musk's X over Grok AI tool and explicit deepfakes

Experts said regulating AI would remain challenging due to the speed of technological change.

Laurence Murphy, senior lecturer in media technology at the University of Salford, said current legal frameworks are struggling to keep pace with rapidly evolving AI systems and stressed that accountability from companies is essential to ensure safeguards are effective.

“The updated approach to online safety laws, particularly around artificial intelligence, shows how complex and fast changing this area is. Legal safeguards need to ensure fair and safe use, but AI evolves rapidly, making oversight difficult," Murphy said. 

He added that legal restrictions alone may not be enough to protect children from online harms as AI systems continue to develop.

"Accountability from companies is key, and while legal restrictions are important, they may not be the only solution to protect children and the public.” 

Copied