UK Police Chiefs Call for Blocking Unsafe Social Media Platforms for Under-16s

The National Police Chiefs' Council and the National Crime Agency issued a joint position on 22 May 2026 declaring that "the online environment, as it currently exists, is not safe for children under 16" [1]. The statement landed on the final days of the UK government's "Growing up in the online world" consultation, which opened on 2 March 2026 and closes on 26 May [2]. Ministers are weighing a statutory minimum age for social media access — with options at 13, 14, 15, or 16 — alongside restrictions on addictive design features, tougher age verification, and new obligations for AI chatbots [2].

The police intervention sharpens a debate that has been building for years. But the central question remains unresolved: can any blocking mechanism actually protect children without creating new harms, and would it survive legal challenge?

The Scale of the Problem

UK police forces recorded 7,062 offences of sexual communication with a child in 2023/24, up 89% from 3,738 in 2017/18 [3]. The NPCC and NCA report that approximately 840,000 adults in the UK pose a sexual risk to children, and in 2025 alone, platforms made nearly 100,000 child sexual abuse referrals to the NCA [1]. Police arrest roughly 1,000 suspected offenders and safeguard around 1,200 children every month [1].

Source: NSPCC / Police recorded crime

Data as of Sep 1, 2024CSV

Sextortion — where offenders coerce children into producing sexual imagery and then threaten to distribute it — has surged. The Internet Watch Foundation confirmed 176 sextortion reports in 2023, up from just 21 in 2022. In the first half of 2025, 153 cases were confirmed, compared to 89 in the same period a year earlier [4]. Boys aged 14–17 are the most targeted demographic, though reports involving girls and children as young as 11 have increased [4].

Source: Internet Watch Foundation

Data as of Jun 30, 2025CSV

When platforms are identified in grooming offences, Snapchat dominates: it was linked to 48% of cases in 2023/24 where the communication method was known, followed by Instagram at 12%, WhatsApp at 8%, and Facebook at 7% [3].

Source: NSPCC / Police recorded crime

Data as of Sep 1, 2024CSV

What the Police Are Actually Proposing

The NPCC/NCA position is more nuanced than headlines suggest. Rather than calling for a blanket ban by app name, the statement targets specific design features that "enable criminals to harm children" [1]. Six categories of risk are identified:

• Mass discoverability: algorithms that surface children's profiles to large numbers of unknown users
• Unrestricted adult-to-child contact: open direct messaging from any adult to any child
• Private encrypted messaging: conversations hidden from any form of oversight
• Algorithmic recommendation of harmful content: systems that promote predatory contacts or dangerous material
• Nude image sharing and livestreaming: features that facilitate coercion into producing sexual content
• Weak age verification: systems easily defeated by entering a false date of birth [1]

Platforms named as containing these features include Snapchat, Instagram, Discord, TikTok, Telegram, Kik, X, and Roblox [1]. The proposal envisions that platforms unable or unwilling to remove these features for child users would face restrictions — though the statement does not specify the exact enforcement mechanism, such as app store removal or ISP-level blocking.

The NSPCC has taken a complementary position. CEO Chris Sherwood argued that unless the government acts, "a social media ban for under 16s could be better than the status quo," and laid out three demands: enforce existing under-13 age limits, stop platforms from using addictive design patterns on teenagers, and block harmful content at the source [5].

Australia's Cautionary Example

Australia provides the closest real-world precedent. Its Online Safety Amendment (Social Media Minimum Age) Act 2024 took effect on 10 December 2025, requiring platforms including Facebook, Instagram, Snapchat, TikTok, X, Reddit, and YouTube to prevent under-16s from creating accounts [6].

The early results are sobering. Local reports indicated that many children bypassed the ban within days. Age-estimation technology was fooled by simple measures — some young users reportedly drew on facial hair to trick facial analysis systems [6]. VPNs proved effective workarounds. A survey found that while 67% of respondents believed the ban would not achieve its aims, only 25% thought it would work [6]. By February 2026, teenagers under 16 were still able to access some platforms [6].

Amnesty International's technology division called the Australian law an "ineffective quick fix," arguing that "the most effective way to protect children and young people online is by protecting all social media users through better regulation, stronger data protection laws and better platform design" [7].

The UK government has been watching Australia closely. The consultation document explicitly references the Australian experience, though ministers have not committed to replicating it.

Which Platforms Would Be 'Unsafe' — and Who Decides?

One of the sharpest criticisms of the police chiefs' proposal concerns the designation process. The NPCC/NCA framework targets design features rather than specific apps, but it does not identify an independent body to assess which platforms meet the threshold, nor does it propose measurable, objective criteria for the "unsafe" designation [1].

The government consultation floats a role for Ofcom, which already regulates online safety under the Online Safety Act 2023. But Ofcom's existing remit is already vast and expanding. The regulator's online safety budget reached £92 million in FY 2025/26, up from £71 million the previous year [8]. Its fee regime — designed to make the tech industry fund its own regulation — only became operational in late 2025, with a notification window for platforms to submit revenue data running into 2026/27 [9].

As of October 2025, Ofcom had launched five enforcement programmes and opened 21 investigations under the Online Safety Act. By December 2025, 76 sites were under investigation [10]. Adding platform-by-platform "unsafe" assessments to this workload would require significant additional resources, and no budget estimate for this expanded function has been published.

Google's UK leadership has warned that an under-16 ban could "drive children towards more dangerous corners of the internet" — unregulated platforms, encrypted messaging apps, and sites beyond the reach of UK regulation [11].

The Evidence Debate: Do Bans Reduce Harm?

The strongest academic challenge to population-level social media restrictions comes from the Oxford Internet Institute. Professor Andrew Przybylski's research, published in the Proceedings of the National Academy of Sciences, analysed large-scale representative panel data and found that social media's effect on adolescent life satisfaction is "nuanced, small at best, reciprocal over time, gender specific, and contingent on analytic methods" [12]. The study did not find that social media use is, in itself, a strong predictor of reduced well-being across the adolescent population.

Przybylski submitted written evidence to Parliament arguing that policy should be based on rigorous evidence rather than assumption, and that the effect sizes found in the highest-quality studies are too small to justify broad restrictions [13].

Against this, the NPCC/NCA position focuses not on general well-being effects but on specific criminal exploitation. The 840,000 figure for adults posing a sexual risk, and the nearly 100,000 annual platform referrals, describe a threat that exists independently of whether average well-being scores shift measurably [1]. The policy question is whether the response should be a population-level block or targeted enforcement against criminal users and non-compliant platforms.

The NSPCC has pointed to a distinct category of harm: addictive design features like infinite scrolling, autoplay, and notification bombardment that are engineered to maximise engagement time, particularly among younger users [5]. Whether these constitute a public health issue comparable to, say, tobacco advertising — justifying a similar regulatory approach — remains contested.

Who Gets Hurt by the Ban Itself?

An estimated 98% of UK 13–17-year-olds own smartphones, and 97% have social media profiles [11]. Among children under 13, 40% already have a social media profile despite existing age restrictions, and among 8–9-year-olds, Instagram profile ownership rose from 8% to 14% in a single year [14]. Ofcom's 2025 report found that even 37% of parents of 3–5-year-olds reported their children using at least one social media app [14].

For LGBTQ+ young people, the stakes are distinct. Research compiled in a systematic review published in the Journal of Medical Internet Research found that social media supports the mental health and well-being of LGBTQ+ youth through peer connection, identity management, and social support [15]. The Trevor Project found that 68% of LGBTQ+ youth said online spaces were affirming, compared to just 38% for home and 16% for community events [16]. In the UK and Australia, 59% of queer young Australians reported receiving some form of mental health support through social media, and 44% join or follow social media groups specifically for queer people [17].

Monash University researchers have argued that the Australian ban "abandons LGBTIQA+ and marginalised youth" by cutting off a primary source of support without providing alternatives [18]. A 2023 Trevor Project survey found that 56% of LGBTQ+ youth seeking mental health care were unable to receive it through conventional channels [16].

Children who rely on social media to maintain contact with family abroad — a significant population in a country where 14% of residents were born outside the UK — would also be affected, though messaging-only platforms like WhatsApp might fall outside the scope of restrictions depending on how "social media" is defined.

Legal Obstacles

Any UK legislation restricting children's access to social media must be compatible with the European Convention on Human Rights, incorporated into domestic law via the Human Rights Act 1998. Two rights are directly engaged:

Article 10 (freedom of expression) protects the right to receive and impart information. Restrictions must be "prescribed by law" and "necessary in a democratic society" — a proportionality test that requires the government to demonstrate that less intrusive measures cannot achieve the same objective [19]. If enforcement of the existing Online Safety Act's children's safety duties has not been fully attempted, a court could find that a blanket block fails this test.

Article 8 (right to private and family life) protects both children's privacy and parents' right to direct their children's upbringing. A statutory ban that overrides parental judgement — preventing a parent from allowing their 15-year-old to use Instagram, for example — raises questions about whether the state is intervening in family life more than is necessary [19].

Mishcon de Reya, a law firm that has published analysis of the proposal, noted that courts are "likely to ask whether less restrictive measures — such as stricter enforcement of existing age limits or mandatory child-friendly design — could achieve similar outcomes" before upholding a platform block [20]. In Australia, two High Court challenges are already proceeding: Reddit argues it should be exempt as an adult-oriented discussion forum, and the Digital Freedom Project contends the law burdens implied freedom of political communication [6].

No published opinion from a King's Counsel or academic constitutional lawyer in the UK has formally assessed whether the proposed block would survive judicial review. The House of Commons Library research briefing on the proposals flags the legal risks but does not offer a definitive view [19].

The Enforcement Gap

Even if the legal obstacles are cleared, enforcement presents formidable challenges. The Online Safety Act's existing age-assurance regime — which requires platforms hosting pornographic content to implement "highly effective" age verification — has already demonstrated the difficulties. By December 2025, Ofcom had fined AVS Group £1 million and 4chan £20,000 for non-compliance, but 76 sites remained under investigation with enforcement far from complete [10].

Extending age assurance to mainstream social media platforms used by tens of millions of UK children would be a substantially larger undertaking. No public body has published an estimated annual cost. Ofcom's online safety budget of £92 million already covers a broad mandate [8], and the fee regime that shifts costs to industry applies only to platforms with qualifying worldwide revenue above £250 million [9] — leaving smaller platforms and emerging services uncovered.

VPNs remain the most obvious circumvention method. They are legal in the UK, widely available, and already used by a significant minority of children. The Australian experience suggests that the mere availability of VPNs is enough to undermine a ban's effectiveness for any moderately tech-literate teenager [6].

Self-Regulation vs. Structural Failure

A critical question is whether the push for blocking represents a failure of platform self-regulation, insufficient enforcement of existing law, or both.

The Online Safety Act 2023 already imposes children's safety duties on platforms. Ofcom's Phase 1 enforcement — focused on illegal content — began in 2025, with Phase 2 covering children's access to harmful content following [10]. But implementation has been slow. The children's safety codes of practice were still being finalised as of early 2026, and full enforcement of children's duties is not expected until later in the year [21].

The NPCC/NCA's own framing implicitly acknowledges this. Their statement argues that feature-based restrictions would "create immediate commercial incentives for platforms to remove dangerous design elements" — something that existing regulation has so far failed to achieve [1]. The question is whether that failure reflects structural loopholes in the law, deliberate non-compliance by platforms, or simply that Ofcom has not yet had enough time and resources to enforce a statute that only received Royal Assent in October 2023.

Blocking platforms accessible via VPN does not address the root cause — platforms that profit from features enabling exploitation. It may instead reduce pressure on platforms to redesign, since blocked services have less incentive to comply with UK regulation they are already excluded from.

What Happens Next

The government consultation closes on 26 May 2026, with a formal response expected in summer 2026 [2]. The options on the table range from a full under-16 ban to narrower measures targeting specific design features, higher age-of-consent thresholds, and mandatory age assurance.

The NPCC/NCA position, the NSPCC's three demands, Australia's troubled rollout, the unresolved legal questions, and the evidence base that remains contested — all of these will feed into a decision that affects millions of children and families. The government has signalled it wants to act. Whether it can design a mechanism that is technically effective, legally durable, and does not create new harms in the process remains the central challenge.