The Policy Case, Charitably Stated
The strongest argument for the UK's new social media ban is not parental anxiety — it is platform failure under existing law. Ofcom's March 2026 enforcement bulletin found that 72% of children aged 8–12 are still accessing platforms that set 13 as their minimum age, despite age restriction policies already on the books. The regulator has already fined companies including 8579 LLC (£1.35 million) and Kick Online Entertainment (£800,000) for failing to implement adequate age checks on adult content sites. The larger, better-resourced social media companies have faced slower and murkier scrutiny.
Prime Minister Starmer's June 15, 2026 announcement responds to this failure directly. The ban covers TikTok, YouTube, Instagram, Facebook, Snapchat, and X — the six platforms where under-13s are already prohibited and under-16s predominate in practice. Messaging services such as WhatsApp and Signal are exempt. For 16–17-year-olds, livestreaming and stranger contact will be disabled by default. AI 'romantic companion' chatbots — a feature proliferating across social platforms — will be restricted to users 18 and over. Regulations are due before Parliament by the end of 2026, with Ofcom enforcement beginning in Spring 2027.
An Honest Reading of the Evidence
Before assessing whether this is the right remedy, the harms deserve clear acknowledgment. Ofcom's 2025 Children and Parents: Media Use and Attitudes report found that 81% of 10–12-year-olds already use at least one social media platform. Three in four children aged 9–17 encounter some form of harm online. Nine in ten parents support the ban — not as ideological reflex, but following a national consultation described as one of the largest the current government has conducted.
The Online Safety Act 2023 already equipped Ofcom with substantial powers: fines of up to 10% of qualifying worldwide revenue or £18 million — whichever is greater — for platforms that fail to protect children. Child safety duties took effect in July 2025. The honest question the June 15 announcement implicitly answers is: those powers exist, and the problem has not been fixed. That failure strengthens the government's political case for escalation.
The Age Verification Problem
Here is where the design of the ban diverges from its ambition. To block under-16s from specific platforms, every user must prove they are 16 or older. Ofcom has been tasked with conducting a 'rapid study' on what constitutes Highly Effective Age Assurance (HEAA), with methods under consideration including photo ID, credit card checks, and biometric facial scans.
The government's own fact sheet specifies that age verification must occur 'without collecting or storing personal data, unless absolutely necessary.' That caveat does significant work. No technically credible age verification system at scale avoids some form of identity data collection. In 2025, 438 security and privacy researchers across 32 countries signed a joint statement calling for an age-assurance moratorium, citing circumvention risks and the invasiveness of the required infrastructure.
The Open Rights Group describes what it calls a 'policy ratchet': each successive child-safety measure — default ISP filters, then the Online Safety Act, now a categorical ban — has delivered less than promised, and each failure has prompted a more extreme next step. The burden falls on users through privacy and speech costs, while the engine of harm — algorithmic amplification and the attention-based business model — remains structurally untouched.
"Once a company holds your iris pattern, you can never be anonymous to it again — biometric data cannot be reset like a password." — DefendDigitalMe
Australia's Six-Month Warning
The UK government has explicitly cited Australia as its model. The Online Safety Amendment (Social Media Minimum Age) Act 2024 came into force on 10 December 2025, making Australia the first jurisdiction to implement a hard age-based platform ban. Six months in, eSafety launched formal compliance investigations into Facebook, Instagram, Snapchat, TikTok, and YouTube on 31 March 2026 — with fines of up to A$49.5 million available for violations.
The early data is sobering. Approximately 70% of children who held accounts before the ban had retained them months later, primarily because platforms had not yet implemented effective age checks. Children losing access to major platforms largely migrated to Roblox, Discord, and other services not covered by the ban. One in two affected children is now seeing less news — a civic information cost that sits alongside the intended harm reduction.
Australia's experience does not prove the policy wrong. It proves it is harder than announced: platform substitution is rapid, the definition of 'social media' is contested and unstable, and enforcement takes years to bite even when political will is present.
What Proportionate Regulation Would Add
A proportionate intervention targets the specific mechanism of harm with the minimum necessary burden on expression and access. The Online Safety Act 2023 already contained most of the right instruments — default-off stranger contact, restricted algorithmic amplification for minors, prohibited promotion of self-harm content — and Ofcom has the power to fine non-compliant platforms heavily.
Proportionate additions would include mandatory algorithmic transparency audits for content shown to under-18s, Ofcom powers to compel changes to recommender systems rather than simply to platform categories, and independent research access so the effectiveness of each measure can be evaluated publicly. These do not produce a clean press-release headline. They address the actual mechanism — the recommendation engine — rather than the platform wrapper.
The question Ofcom's 2026 age verification study must answer is not only how to verify age, but what the privacy cost of doing so is, and who bears it. Until that answer exists, Spring 2027 enforcement will rest on infrastructure that has not yet been demonstrated to work at scale.