The UK government is preparing an “Australia plus” plan to keep children under 16 off major social media platforms, while also pushing new limits into gaming apps, AI chatbots, livestreaming, disappearing messages, and late-night scrolling for older teenagers.
Prime Minister Keir Starmer is expected to outline the package on Monday, according to The Guardian, after months of consultation over children’s use of phones, social platforms, games, and AI companions. The government has not yet published the final platform list or the full enforcement model, so the most important details are still pending. What is already clear is that the proposal is no longer only about whether a 14-year-old can open Instagram. It is about which parts of the internet will have to know a user’s age before letting them chat, scroll, stream, message, or use AI companions.
The reported plan would ban under-16s from major social platforms in a model similar to Australia’s live age-restriction regime. It would also go further by targeting features on services that may not be covered by a direct social-media ban. Gaming apps, for example, could be required to remove or restrict options that let young users talk to adult strangers. Under-18s would reportedly be blocked from romantic or sexual AI chatbots, and older teens could face rules meant to prevent scrolling after 8:30 p.m.
Why the UK plan is broader than a social app ban
The technology issue is the scope. A simple under-16 account ban would mostly focus on large social networks. The UK plan appears to treat risky functions as portable: a livestream can exist on a social app, a gaming platform, or a creator tool; stranger chat can sit inside a game as easily as inside a messaging app; an AI companion can be marketed as entertainment, tutoring, role-play, or emotional support.
That approach follows the language of the UK government’s own “Growing up in the online world” consultation, which asked about minimum ages for social media, risky design features such as infinite scrolling and autoplay, digital age of consent, age assurance, and school phone rules. A March government announcement framed the consultation as covering social media, gaming platforms, and AI chatbots rather than treating child online safety as a problem contained inside a small set of apps.
The House of Commons Library’s June briefing says Part 3 of the Children’s Wellbeing and Schools Act 2026 will require the government to impose some form of age or functionality restrictions for children under 16. That matters because the legal mechanism may not be limited to a single blacklist of social networks. It could become a framework for regulating features that migrate across platforms.
Age checks become the hard part
The practical burden will fall on age assurance. Platforms cannot enforce a child-access rule without some way to determine whether a user is under 16, under 18, or an adult. The UK already has Online Safety Act duties that require regulated services to assess child access and protect children from harmful content, but a broad social-media age rule would push age checks closer to the front door of mainstream services.
Ofcom has been moving in that direction. In March, the regulator told major platforms including Facebook, Instagram, Roblox, Snapchat, TikTok, and YouTube to enforce minimum ages, address grooming risks, make feeds safer, and test products more rigorously. The same notice said popular sites and apps would need highly effective age checks, not just terms-of-service language that children can ignore.
That is where the policy becomes difficult for privacy and product design. Age assurance can include document checks, facial age estimation, payment-card signals, mobile-provider checks, account history, or third-party identity services. Each method creates tradeoffs: stronger checks may block more underage accounts, but they can also collect sensitive data, exclude people without suitable documents, or normalize identity checks across services that previously allowed pseudonymous use.
Critics have already warned about that risk. Tech Policy Press summarized the concern this week: if a platform must prove whether a child is present, it may end up verifying everyone. That could expand surveillance, reduce anonymity, and create new breach targets even when the policy goal is child safety.
Australia is the model, but not a finished answer
Australia’s system is the live comparison point. Its social media minimum-age rules took effect on December 10, 2025, and the country’s eSafety Commissioner says age-restricted platforms must take reasonable steps to prevent under-16s from creating or keeping accounts. The regulator identifies Facebook, Instagram, Snapchat, Threads, TikTok, Twitch, X, YouTube, Kick, and Reddit as age-restricted platforms, while noting that online gaming and standalone messaging apps are excluded under its current rules unless they include social-media-style features.
The Australian regime is not framed as a punishment for children or parents. The compliance duty sits on platforms. Civil penalties can reach up to 49.5 million Australian dollars for companies that fail to take reasonable steps, according to the eSafety Commissioner’s guidance.
Australia also shows why the UK wants to go beyond a platform list. eSafety says platforms removed access to 4.7 million under-16 accounts by mid-December 2025, but the regulator’s March 2026 update also emphasized that implementation takes time and that age restrictions are only one part of a wider safety system. Children can move to less visible services, use shared devices, misstate ages, or find workarounds. Harmful design patterns can also reappear in features that sit outside the narrow legal definition of social media.
Games and AI companions are now part of the same fight
The most consequential part of the UK approach may be its treatment of platforms that do not look like traditional social networks. Roblox, Discord-like communities, game voice chat, livestreaming, and AI companion apps all mix entertainment with social interaction. For a child, the relevant risk may not be the category label on an app store page. It may be who can contact them, what the recommendation system amplifies, whether messages disappear, whether strangers can move a conversation elsewhere, or whether an AI chatbot simulates intimacy.
That makes enforcement more complex. A game could be low risk in solo mode and high risk in open chat. A messaging feature could be benign among verified friends and risky when adults can contact minors. An AI chatbot could be an ordinary homework helper in one setting and a sexualized companion in another. Rules written around app categories will struggle to keep up with products that keep blending social, gaming, creator, and AI functions.
For platforms, the direction of travel is still visible. Services likely to be used by children will need clearer age gates, safer defaults, feature-level controls, audit trails for risk decisions, and evidence that they are not relying on self-declared birthdays. Gaming and chatbot companies that once saw online safety as a social-network problem should assume that regulators now see their products as part of the same child-safety map.
What to watch next
The immediate question is how precise Starmer’s announcement will be. A strong policy signal is not the same as an enforceable rule. The platform list, start date, appeals process, technical standards for age assurance, treatment of VPNs, rules for games and messaging services, privacy safeguards, and penalties will decide whether the UK plan becomes a workable child-safety system or a broad mandate that is difficult to enforce without checking the identity of ordinary users.
The UK is not moving in isolation. Australia has already made under-16 social media restrictions operational. The EU, U.S. states, and other governments are also tightening rules around minors, addictive design, age-appropriate experiences, and AI companions. The UK plan is notable because it appears to combine all of those tensions in one package: social media access, platform design, gaming chat, AI intimacy, youth sleep, identity checks, and regulator oversight.
If the details match the early reporting, the UK will be testing whether child online safety can be regulated by feature and function rather than by app name. That is the part every platform should be watching.
