UK to ban under-16s from social media by spring 2027

Keir Starmer's government has decided that British children under 16 should not be on social media, and it wants the law on the books before Christmas. Enforcement is pencilled in for spring 2027. TikTok, Facebook, Instagram, Snapchat, X and YouTube are all caught by the proposal. WhatsApp and other messaging apps are not.

The blueprint is Australia's, which switched on last December, with a few extra bolts added. Livestreaming and stranger contact with under-16s will be blocked across a wider set of services, gaming platforms included. AI companion chatbots and any intimate chatbot features will be off-limits to anyone under 18. Overnight curfews and forced breaks in infinite-scroll feeds are also being floated, with detail promised next month.

The bit nobody has figured out

Every conversation about banning kids from social media eventually arrives at the same unanswered question: how do you actually know someone is 16?

Australia's own regulator has already given the world a preview. More than two-thirds of Australian kids who kept their accounts after the ban did so because the platforms simply never asked them to prove their age. Not a clever workaround. Not a VPN. The systems just shrugged and let them in.

The UK is leaning on Ofcom to come up with what it calls "highly effective age assurance", with a plan due in October. What that means in practice is the interesting part. Government-issued ID uploads? Face-scanning estimators? Bank-card checks? Each option carries a privacy cost for every adult who has to prove they aren't 15 before scrolling.

The politics

Reaction has split in predictable ways.

• Amnesty International UK called it the right diagnosis with the wrong prescription. Their argument is that the platforms are the problem and should be fixed, rather than the kids being shown the door.
• The IAPP's Joe Jonas pointed out the obvious timing issue: Starmer's government may not be in office long enough to see spring 2027.
• Nine in ten of the 116,000 parents the government surveyed back the policy, so the public appetite is not in dispute.

Public mood and workable policy are different animals though. The Online Safety Act is already in force and platforms are already supposed to be keeping children away from the worst content. A blanket age cut-off is a much louder intervention, and it depends entirely on a verification regime that no country has yet built convincingly.

What to watch

Two dates matter. October, when Ofcom publishes its age-assurance plan, will tell us whether the government has a credible technical answer or a polite hope. The legislation itself, expected before Christmas, will show how much of the harder material, the curfews, the chatbot restrictions, the gaming scope, survives contact with the lobbyists.

Between now and spring 2027, the platforms have eighteen months to either build something that genuinely works or to keep doing what Australia found them doing, which is not very much at all.