French parliament bans social media for under-15s

When French schools reopen in September, students under 15 may find themselves on the wrong side of a new law. On 20 July, a joint committee of deputies and senators agreed on a ban preventing under-15s from holding social media accounts (Le Progrès, Midi Libre). If both chambers adopt the text, France becomes one of Europe's clearest test cases for a question every parent has asked: can you actually keep teenagers off Instagram?

Parliament chose the broad approach. Rather than blacklisting only the most harmful platforms and letting parents consent for others (the Senate's preference), lawmakers drew a single age boundary: no social media before 15, with exceptions for educational services like Wikipedia (Le Parisien, LCP). In other words, France bet on one enforceable line instead of case-by-case platform policing.

How do you check age without checking identity?

Passing a ban is the easy part. Making it work without turning every login into an ID checkpoint is harder. France is not proposing that TikTok demand a teenager's passport. The model its backers describe uses a third-party verifier, an independent service that checks whether someone is above or below 15 and passes only a yes-or-no result to the platform (RTL).

Think of a bouncer at a club entrance. The bouncer sees your ID, confirms you're old enough, and nods you through. The bar inside never sees your name or birth date. In cryptography, this idea has a name: a zero-knowledge proof, a way to prove one fact ("old enough") without handing over the evidence behind it. The EU's expert panel on child safety recommends this as the privacy-respecting path forward (Commission panel). The EU is also building a Digital Identity Wallet designed to let citizens prove things like their age without revealing everything else on their ID (EU Digital Identity Wallet).

This is not theoretical. Ofcom's first review of age assurance in the UK found 69 million completed age checks across 32 services, using methods from facial estimation to photo-ID matching (Ofcom). But none of these systems have been deployed at the scale social media demands.

The gap between a gate and a wall

An age gate on a pornography site has one clear entrance. Social media is different. There are existing accounts, shared family tablets, older siblings' phones, VPNs, and a population of 14-year-olds with strong motivation to get around any barrier.

The Commission's expert panel warns that early experience with blanket bans suggests many minors still find ways through (Commission panel). The simplest check, the familiar birthday dropdown, is not considered effective by Ofcom (Ofcom). Facial age estimation sits in between, but it becomes least accurate right around 14–15, where the legal line falls.

France writes the rule. Who enforces it?

France can pass a law, but Meta, TikTok, and Google run their EU operations from Ireland. Enforcement therefore travels through the EU's Digital Services Act, which gives the European Commission and Ireland's national regulator shared oversight of very large platforms (DSA). The Commission is already scrutinising Meta and TikTok over design features that may hook younger users, with possible fines of up to 6% of global turnover (Slaughter and May, DSA). Spain is building its own framework combining age thresholds with platform design duties (20minutos). Ireland's Michael McGrath has argued the complexity calls for EU-wide rules rather than a patchwork of national ones (Irish Examiner).

France is trying to force platforms, verifiers, and EU regulators to solve the plumbing after the law is passed. Supporters want it working by September's rentrée (RTL), but passing a law and deploying age-verification systems across every platform are very different timelines. What matters next: whether platforms build privacy-preserving checks, whether teenagers find ways around them, and whether EU regulators treat enforcement as urgent or let it drift.