If you're tired of censorship and dystopian threats against civil liberties, subscribe to Reclaim The Net. French lawmakers spent Monday afternoon behind closed doors agreeing that children under 15 should be locked out of social media. They built a legal requirement for every adult in France to prove who they are before they can post anything. Seven deputies and seven senators, sitting as a commission mixte paritaire, took roughly two hours to reconcile two very different versions of Renaissance deputy Laure Miller’s bill. The text they produced follows the National Assembly almost exactly. A new provision slotted into France’s 2004 digital economy law declares that “access to an online social network service provided by an online platform is forbidden to minors under fifteen years of age.” Both chambers voted Tuesday on the committee’s conclusions. The Senate signed off first, the National Assembly followed 279 to 81. Support ran from the Rassemblement National through the centrist bloc. La France Insoumise voted against, and the Socialists abstained. Emmanuel Macron celebrated within hours. “I had committed to it, and now it’s been voted on: social media will be banned for those under 15 starting back-to-school,” he wrote, thanking the parliamentarians. “It’s up to the Constitutional Council to rule, then on to action to make this measure concrete and protect our children online.” The Council now has a month to weigh the text against the constitution. If it survives, Macron signs, a formality given how hard he pushed for it. The arithmetic of an age ban has not changed since the first one was drafted. Blocking 14-year-olds means screening everyone, because a platform cannot tell who is 14 without asking who all of us are. Age assurance, whatever the marketing calls it, runs on identity documents or a state-issued digital identity. You photograph your passport, or you sign in through France Connect. The other option on the table, facial age estimation, is unreliable enough that a British study found a drawn-on mustache can fool it, and the CNIL has already refused to allow the technique in tobacco shops. French officials answer all this with the phrase “double-anonymat,” a term ministers and regulators have leaned on since 2023 to describe an arrangement where a third party inspects your ID and hands the platform a signed token instead of your name. AI Forensics found in 2025 that AgeGo, a verification vendor working for numerous pornography sites, was harvesting the full URL of the video each user wanted to watch. The committee then stripped out the guardrails. Earlier versions of the bill instructed providers to use technical solutions conforming to a standard drawn up by Arcom after consultation with the CNIL, and gave Arcom the power to send formal notices to companies that ignored it. The compromise deletes the standard, deletes the enforcement, and deletes the CNIL’s advisory role over age verification systems. France is about to demand age verification without defining what age verification means, and without any regulator supervising the firms that will end up holding millions of identity documents. Platforms pick their own vendors and users hand their passports to whoever bids lowest. It was Brussels that drove the deletions. The European Commission raised objections two weeks ago, worried about the amount of discretion the Senate version handed Arcom over which platforms would be blacklisted. Parliamentarians responded by scrubbing Arcom from the law entirely, including its role in checking whether the law is being followed. Miller told Public Sénat that maintaining a blacklist risked non-compliance with the Digital Services Act, and that the Commission confirmed as much in informal exchanges, “Hence the idea of returning to a general ban, more comprehensible and without additional constraints for the platforms.” A general ban, then, chosen because it asks less of the companies and more of the public. Senate rapporteur Catherine Morin-Desailly, who had championed the abandoned two-tier approach, downplayed the retreat. “As it stands, messaging services are excluded from the definition of social networks. When you announce a general ban, there are already exceptions,” she said, noting that messaging services fall outside the definition. She was franker about the law’s reach, saying its scope is limited because it does little beyond setting a digital age of majority while European rules are about to be written anyway. Laurent Lafon, the centrist senator who chairs the Senate culture committee and sat on the joint committee, conceded that “Everyone is aware of the limits of the text, both legally and in terms of enforcement,” while praising the message sent to platforms. There are some carve-outs but they’re thin. Online encyclopedias survive, along with educational and scientific directories and platforms for developing and sharing open-source software and open-source educational projects. Nothing exempts small or non-commercial services, which means a Mastodon instance run by three volunteers falls under the same obligation as Instagram. The people building the decentralized, non-extractive corners of the internet get handed a compliance burden designed for Meta, and the surveillance business model that created the harm goes untouched. The law takes effect on September 1, with accounts created before that date getting a four-month grace period. Covered platforms have to block under-15s from opening new accounts from that day, and close the accounts those users already hold by January 1, 2027. Enforcement is where these laws tend to fall over. Australia banned under-16s from social media in late 2025, and its own regulator has acknowledged that two thirds of the affected teenagers get around it. Article 28 of the DSA, the provision the Commission and Paris both lean on, states in its third paragraph that it does not require platforms to process additional personal data to work out whether a user is a minor. An age verification mandate does exactly the thing the text says it does not require. Other routes existed and were not taken. Mandatory interoperability would let people walk away from a platform without abandoning everyone they know on it. Banning surveillance advertising would defund the infinite scroll. Rules on recommendation systems would go after the design choices that produced the harm French lawmakers keep citing. What parliament chose instead leaves the business model untouched and bills the public for the fix, payable in identity documents. If you're tired of censorship and dystopian threats against civil liberties, subscribe to Reclaim The Net. The post French Parliament Reaches Deal on Social Media Ban appeared first on Reclaim The Net: Free Speech, Privacy, Digital Rights.