The UK government plans a new default safety regime for 16- and 17-year-olds on social media. In its GOV.UK announcement, ministers said platforms would be expected to switch on overnight curfews from midnight to 6am by default and disable features linked to prolonged use, including autoplay and continuously personalised feeds. The government says the first regulations will be laid before Parliament by the end of 2026, with the measures expected to take effect in spring 2027. That timetable places the changes alongside the planned ban on social media services for under-16s, and it signals that ministers want continuity in child protection rather than a sudden reduction in safeguards at 16.
The structure of the package matters. The announcement describes a default setting rather than an absolute prohibition for older teenagers. Sixteen- and seventeen-year-olds would retain the ability to change their own settings, but the starting position would be a more restrictive service. That gives ministers a middle course between parental expectations and adolescent independence. For regulators and platforms, the policy is as much about product design as content. The government is targeting recommendation systems, autoplay and other features that extend session length, not only unlawful or harmful posts.
In practical terms, the proposal would require social media companies to build age-based product experiences that differ for under-16s, for 16- and 17-year-olds, and for adults. The GOV.UK statement points to two priority areas: overnight access and attention-grabbing design. Platforms would need reliable age assurance, time-based controls, auditable default settings and a straightforward route for older teenagers to alter those settings if they choose. The accompanying note on enforcement is brief, but the direction is clear. Ministers say there will be robust implementation and enforcement once the rules are in force. That is likely to place pressure on firms to show that curfews, autoplay settings and feed configuration work as required and cannot be easily bypassed.
The government is presenting the plan as evidence-led. According to the pilot research referenced on GOV.UK, more than 300 teenagers and parents across the UK took part in a trial of social media restrictions. Families reported that overnight curfews quickly became routine and that they were associated with better sleep and concentration. That matters because the policy case is being framed around everyday functioning rather than abstract digital risk. Ministers are tying the proposal to sleep, school and college focus, and time with family and friends. The evidence cited in the announcement is qualitative rather than a full national impact study, but it gives the government a clearer basis for claiming public support and operational feasibility.
The wider policy objective is to remove what ministers describe as a cliff edge in protections when a child turns 16. Under the government’s existing plan, social media services would be banned for under-16s from spring 2027. The new package is intended to make the move into later adolescence a controlled step rather than a sudden shift from prohibition to adult-style product design. This is also where the government’s published work on age assurance and circumvention becomes relevant. If children can misstate their age or move easily between services, default protections lose force. The policy therefore rests not only on new rules, but on whether platforms and regulators can make age checks credible without creating new privacy or access problems.
Alongside the social media package, the Technology Secretary intends to bring forward separate measures for AI chatbots used by children. The GOV.UK announcement says under-18s would be prompted to take regular breaks, with the aim of discouraging prolonged or intense use. Ministers are also working with regulators and other departments on services that provide dangerous, misleading or unverified mental health advice. The government says all options remain under consideration, including banning chatbots that pose a serious threat to children. That is a notable change in emphasis. It suggests ministers are beginning to treat some AI services not simply as general consumer tools, but as products that may require child-specific safety rules where they mimic advice, companionship or therapeutic support.
The package also extends beyond platforms and into public guidance and the curriculum. Ministers say the Kids Online Safety Hub will be expanded so that children, parents and guardians have more official guidance on using AI safely and with confidence. That gives the government a public information route alongside formal regulation. Schools are expected to carry part of the delivery burden. From September 2026, RSHE lessons are due to cover critical thinking about new technology, including AI and chatbots, as well as mis- and disinformation, violent material and misogynistic content. From September 2028, the National Curriculum is due to embed media literacy across subjects and strengthen teaching on source analysis, bias, AI, data science and technological bias.
For platforms, the practical effect is that compliance will no longer be limited to removing harmful content after the fact. The government is moving towards a model in which service design, recommendation settings and time-of-use controls are themselves part of the safety case. Firms with large teen user bases will need to prepare for product changes, assurance testing and closer regulatory scrutiny before spring 2027. For parents and schools, the offer is more modest but still significant. The measures would not bar 16- and 17-year-olds from social media altogether, yet they would reset the default environment at the point where many families report a sharp rise in screen time. As presented in the GOV.UK announcement, the policy aim is simple: extend child protection into later teenage years without treating 16-year-olds as full adults online from one day to the next.