Starmer to issue social media curfew for teenagers in bid to crack down on addictive scrolling

The Announcement of Overnight Social Media Curfews

Sir Keir Starmer has announced plans for a voluntary overnight social media curfew aimed at 16- and 17-year-olds as one of his final acts as prime minister. Technology secretary Liz Kendall will set out the measures that introduce default settings to block access between midnight and 6am. The proposals target addictive scrolling through automatic disabling of features such as endless video reels and algorithmic feeds. These steps form part of wider efforts coordinated from No 10 and Whitehall to address youth engagement with platforms. The announcement arrives in recent days ahead of the Makerfield by-election, with the contest winner and incoming prime minister Andy Burnham set to inherit responsibility for resolving outstanding details.

Liz Kendall, the Culture Secretary, outlined the government’s position in detail during the announcement, stating that the consultation had revealed overwhelming support for structured restrictions. She emphasised that parents and teenagers alike had voiced concerns about the mental health impacts of late-night scrolling, with many young people admitting they struggled to switch off without external prompts. Kendall noted that these curfews would form a core element of the wider Online Safety framework, complementing existing duties on platforms to assess and mitigate systemic risks to children rather than relying solely on reactive enforcement.

The feedback gathered during the consultation painted a consistent picture: parents reported feeling powerless against algorithmic feeds designed to maximise engagement, while teenagers described pressure to remain online to avoid social exclusion. Officials concluded that overnight curfews represented a proportionate response that balanced protection with the recognition that older adolescents might require more flexible arrangements. By embedding these measures within the Online Safety Act’s risk-assessment regime, ministers aim to shift responsibility onto platforms to demonstrate how their design choices actively support healthy usage patterns, rather than treating curfews as an isolated policy fix.

The Broader Regulatory Context

The trajectory of UK social media regulation has been marked by incremental yet often faltering steps, culminating in the Online Safety Act 2023 which sought to impose duties of care on platforms regarding harmful content. Earlier efforts, including the 2017 Green Paper on internet safety and subsequent consultations under successive Conservative administrations, repeatedly stalled amid industry lobbying and concerns over free speech. Starmer's imposition of a curfew on 16- and 17-year-olds represents a decisive pivot from the Act's emphasis on risk assessment towards prescriptive age-based restrictions, echoing but exceeding measures in the EU's Digital Services Act and Australia's eSafety framework, where default protections have been trialled with varying degrees of enforcement success. Politically, this move appears calibrated as one of Starmer's final acts to cement a legacy on child protection before any transition, leveraging public anxiety over youth mental health to neutralise potential criticism from both progressive and conservative flanks.

Technical Details of Default Settings

Default settings will block access for 16- and 17-year-olds between midnight and 6am if the proposals advance. Algorithmic feeds and endless video reels will face automatic disablement under the same framework. Technology secretary Liz Kendall holds responsibility for presenting these elements to Parliament and industry stakeholders. Implementation depends on platform compliance with the default configurations. The government has not yet confirmed enforcement mechanisms beyond the initial defaults. These provisions connect to existing regulatory discussions in Westminster about digital design standards.

Age Verification and Implementation Challenges

Age verification technologies underpinning these default settings rely heavily on biometric or document-based checks, yet their reliability remains contested, particularly when applied across diverse platforms with differing user interfaces. Andy Lulham of Verifymy has emphasised the inherent difficulties in accurately segmenting users into under-16, 16-17, and adult categories without creating exclusionary barriers or false positives that could inadvertently restrict legitimate access. These challenges are compounded by the need for real-time, cross-device consistency, where VPN usage and shared accounts further erode the precision of any technical solution. Comparisons with Australia's age-verification mandates highlight the risks of overreach, where implementation delays have exposed enforcement gaps, while EU approaches have prioritised systemic risk audits over blanket curfews.

The European Union has adopted a more prescriptive stance through its Digital Services Act and forthcoming AI Act, requiring platforms to implement default privacy-by-design settings for minors and to conduct systemic risk assessments that explicitly consider the addictive potential of algorithmic recommendations. Several member states are now piloting time-limiting features that automatically restrict social media access during school hours and overnight, backed by fines of up to six percent of global turnover for non-compliance. These measures place the onus on companies to prove their systems are safe by design, rather than waiting for evidence of harm to accumulate.

Australia’s age-verification trials, conducted across multiple states, have highlighted significant implementation gaps that the UK would do well to study. Early evaluations revealed inconsistent enforcement, with many platforms relying on weak self-declaration methods that teenagers could easily bypass, while concerns about data privacy and exclusion of vulnerable groups have slowed wider rollout. Policymakers in Westminster are therefore examining whether a centralised, privacy-preserving verification infrastructure, rather than fragmented industry-led solutions, might deliver more reliable outcomes before similar trials are attempted here.

Connection to the Under-16s Ban

Last month the government unveiled a social media ban for under-16s expected from next spring. The ban covers Snapchat, TikTok, YouTube, Instagram, Facebook and X while excluding messaging services such as WhatsApp and Signal. The new curfew for 16- and 17-year-olds extends the same regulatory thread. Both policies emerged from the same department led by Liz Kendall. The under-16s measure sets a precedent for age-based restrictions across major platforms. Details of how the two initiatives will interact remain subject to further clarification by the incoming administration.

Findings from the Government Pilot Programme

Families who took part in a government pilot involving more than 300 teenagers and parents across the UK reported that overnight curfews helped improve sleep and concentration. The programme tested the practical effects of restricted access during night hours. Participants spanned multiple regions and provided feedback on daily routines. These results informed the current proposals announced by Sir Keir Starmer. The pilot data focused on measurable changes in sleep patterns and attention levels.

Additional Provisions on AI Chatbots

The proposals require under-18s to take regular breaks while using AI chatbots. A separate crackdown targets AI services that provide dangerous or unverified mental health advice. These elements sit alongside the social media curfew within the same policy package. Technology secretary Liz Kendall will include the AI requirements in the forthcoming announcement. The measures aim to limit continuous engagement with automated systems. They apply to all under-18s regardless of the social media curfew age bracket.

Under the new provisions, platforms will be required to introduce mandatory breaks after thirty minutes of continuous interaction with AI chatbots, delivered through algorithmic prompts that pause the conversation and suggest alternative activities. These prompts must be non-dismissible for users under eighteen and accompanied by clear explanations of why the break is being enforced. The government has also signalled a broader crackdown on AI mental health services that provide dangerous or unverified advice, with Ofcom empowered to issue immediate takedown notices where content risks exacerbating self-harm or eating disorders.

From September, schools will receive updated statutory guidance requiring them to incorporate digital wellbeing modules that address the specific risks posed by conversational AI. This includes training for teachers on identifying when pupils are over-relying on chatbots for emotional support. Meanwhile, the EU’s approach offers a useful comparator: several member states are trialling default restrictions that prevent children under thirteen from accessing unmoderated AI tools altogether, with age-assurance checks required at the point of first use rather than relying on self-declaration.

Criticism from Opposition and Charities

Conservative shadow education secretary Laura Trott criticised the plans during the initial response. The NSPCC stated that the proposals will not be enough on their own. The 5Rights Foundation argued that curfews only manage exposure to risk without incentivising change in the tech industry. Critics questioned overall effectiveness given the voluntary framework.

Chris Sherwood of the NSPCC described the measures as “sticking plaster” solutions that fail to address the underlying commercial incentives driving harmful design. He argued that without stronger enforcement powers and meaningful financial penalties, platforms would continue to prioritise engagement metrics over child safety, leaving families to navigate an increasingly complex digital environment with limited real protection.

Colette Collins-Walsh from the 5Rights Foundation echoed these concerns, stating that the tech industry is not incentivised to change its fundamental business model. She warned that voluntary commitments and light-touch guidance would prove insufficient against companies whose algorithms are explicitly engineered to maximise time spent online. Andy Lulham of Verifymy added that robust age assurance remains essential, noting that without technically reliable methods to verify users’ ages, even well-intentioned rules around curfews and chatbot breaks would be easily circumvented by determined teenagers.

Challenges for Future Delivery

Key details of the curfew and AI measures now rest with incoming prime minister Andy Burnham. Platform operators must still confirm how defaults will operate across devices used by 16- and 17-year-olds. The exclusion of messaging services from the under-16s ban creates a parallel boundary for the new rules. Local councils and devolved governments have not yet received formal guidance on supporting families. The pilot programme results provide one data point but leave open questions about wider rollout. Delivery timelines remain tied to the post-by-election period. Implementation timelines are expected to stretch into 2026, allowing platforms latitude to develop bespoke protocols, though this delay risks leaving current users unprotected amid rapid AI advancement.

Legacy Implications and Economic Considerations

Cost implications for technology companies are substantial, with estimates suggesting multimillion-pound investments in compliant systems that smaller platforms may struggle to absorb, potentially accelerating market consolidation. The technical hurdle of distinguishing age groups without compromising privacy or usability underscores a broader tension between regulatory ambition and engineering reality, where no single verification method has yet demonstrated scalability across the fragmented digital ecosystem. As one of Starmer's concluding initiatives, the measures carry symbolic weight that could shape perceptions of his premiership's digital legacy. The absence of robust auditing mechanisms raises questions about whether these measures will genuinely mitigate harms or merely offer performative compliance.