The United Kingdom's government has announced that social media platforms operating in the country will be required to default-restrict access for 16 and 17-year-olds between midnight and 6am — a measure framed as a landmark intervention in the mental health crisis among young people. Under the proposal, platforms including Instagram, TikTok, and YouTube would also be required to disable so-called addictive design features by default: auto-play video, infinite scroll, and algorithmic push notifications that have long been central to how these companies extract time and attention from their youngest users.

The announcement comes from the Department for Work and Pensions under Secretary of State Liz Kendall, framed as part of a broader government push to treat youth screen dependency as a public health issue rather than a parenting problem. The political framing is deliberate — it positions the state as intervening on behalf of children against the commercial interests of Silicon Valley platforms, at a moment when that position is broadly popular across the electorate.

Here is the catch that the press release buries: the curfew is opt-out. A 16-year-old who wants to scroll TikTok at 2am needs only to go into their account settings and switch the restriction off. The government has not said it will require platforms to make that process difficult, time-delayed, or subject to parental approval. In practice, this means the policy's entire protective effect rests on the assumption that teenagers in the grip of compulsive late-night scrolling will not bother to disable a setting — an assumption anyone who has ever been a teenager, or raised one, might find optimistic.

The opt-out structure is not an oversight. It is the architecture of a policy designed to be passable. Platforms would almost certainly challenge a mandatory hard curfew through legal and regulatory channels, citing user autonomy and the thorny question of age verification at scale. An opt-out default is the compromise that lets the government claim action while giving the platforms a workable off-ramp. Whether that compromise actually helps teenagers is a separate question from whether it was politically achievable.

The science underlying the concern is real, even if the policy mechanism is weak. A substantial body of peer-reviewed research, including longitudinal studies drawing on data from tens of thousands of adolescents, has linked heavy late-night social media use to disrupted sleep architecture, depressive symptom clusters, and anxiety — with effects that are measurably larger in girls than boys. The specific features targeted by this proposal — infinite scroll and auto-play — are not accidental design choices. They were engineered, with considerable investment, to reduce the friction that would otherwise cause a user to stop and go to bed. Calling them "addictive" is not hyperbole; it is a reasonable description of their intended function.

What is missing from the government's framing is any serious reckoning with enforcement. The Online Safety Act, which passed in 2023, gave Ofcom new powers over platforms operating in the UK, including the ability to levy fines of up to ten percent of global annual turnover. But Ofcom's track record on enforcing age-related protections has been slow, and the regulatory infrastructure for verifying that a platform has correctly applied default settings to a correctly identified age group does not yet robustly exist. The announcement describes what platforms must do. It is considerably less clear about what happens when they do not do it well — or at all.

There is also a displacement question the policy does not address. Social media does not exist only on the apps the government named. A 17-year-old whose Instagram goes dark at midnight can open a browser, visit the same content on the web, and encounter zero restriction. The proposal applies to apps, not the open internet — a gap that reflects either a failure of scope or a deliberate narrowing to avoid a fight the government does not want to have with ISPs and browser vendors.

None of this makes the policy worthless. Default-on restrictions do change behavior at the margins — behavioral economists have documented this effect across dozens of domains, from pension enrollment to organ donation. A teenager who is not a determined circumventer may simply go to sleep because the app is off. For that teenager, the policy works. The honest version of this announcement would say: this will help some young people some of the time, it is not a substitute for platform accountability, and the opt-out clause is a political concession to industry that we are calling a feature. That version does not get written into press releases. It is, nonetheless, what is actually true.