The UK government wants social media platforms to switch on default overnight curfews for 16- and 17-year-olds, a move that would push online safety regulation beyond under-16 bans and into the design of social apps used by older teenagers.
Under the proposal announced by the Department for Science, Innovation and Technology, social media services would set a midnight-to-6am curfew by default for 16- and 17-year-olds. Autoplay, infinite scroll and highly personalised feeds would also be switched off by default, although older teens would still be able to change their settings.
The idea is to avoid a sharp safety cliff when children turn 16. The UK is already preparing a wider under-16 social media restriction from spring 2027, but ministers argue that 16- and 17-year-olds still need protection from the most addictive design features, especially around sleep, concentration and mental health.
The proposal lands in a wider global shift toward teen-safety regulation. Ofcom has already raised questions about whether platforms such as TikTok and YouTube provide enough youth safeguards, while US states have also tested age-based access limits and parental controls. Earlier reporting on TikTok and YouTube youth safeguards and Utah’s teen social media restrictions shows how quickly this has become a mainstream policy debate.
A ban tells users they cannot access a service. A default curfew changes the product environment without fully removing choice. That difference matters. The UK government is trying to make healthier defaults the norm while still leaving older teenagers some agency over their settings.
The policy also targets design rather than only content. Autoplay, infinite scroll and algorithmic feeds are not illegal speech, but they are mechanics that keep people watching, tapping and scrolling for longer. Regulators are increasingly treating those design choices as public-health and child-safety issues.
That is a major change for platforms. Social media companies have long argued that users decide how much time they spend online. Governments are now saying the architecture of the product shapes those decisions, especially for younger people.
The announcement also points toward new rules for AI chatbots used by under-18s. Ministers are considering mandatory breaks, restrictions on chatbots that give dangerous or unverified mental-health advice, and new guidance for children, parents and guardians on safer AI use.
That AI section may become the more important part over time. Teenagers are no longer only scrolling through human-made posts. They are also interacting with AI assistants, character bots and emotionally responsive systems that can give advice, simulate friendship or become part of a daily routine.
A recent TechBooky draft on OpenAI’s reported ChatGPT speaker framed the same issue from a hardware angle: AI is moving from screen-based tools into more personal spaces. If AI systems become companions, tutors or informal advisers, child-safety rules will have to cover more than social feeds.
Platforms will likely push back on several grounds. They may argue that teenagers can already use screen-time tools, that age assurance remains imperfect, and that overly rigid defaults could reduce access to legitimate social connection, education and support.
Those objections are not trivial. A midnight curfew may help some teenagers sleep, but it may also affect young people who rely on social platforms for community, creativity, activism or support outside traditional hours. The government says older teens can change their settings, but easy opt-outs may weaken the policy’s impact.
There is also a technical enforcement challenge. Platforms need reliable age assurance without creating new privacy risks. If companies collect more identity data to comply with safety rules, regulators will have to ensure that the cure does not create a new data-protection problem.
The bigger signal is clear: online safety regulation is moving from content moderation to product design. Governments are no longer asking only whether platforms remove harmful posts. They are asking whether the product itself is designed to keep young people hooked.
For tech companies, that means child-safety compliance will increasingly touch interface design, recommendation systems, AI chatbots, notifications, default settings and age assurance. It will not be enough to publish safety policies after the fact.
The UK proposal may still change before implementation, but it captures the direction of travel. Teen safety is becoming a design obligation, and AI chatbots are about to be pulled into the same regulatory conversation as social media feeds.