EU Kids Act consultation opens as Meta unveils new child safety tools

Brussels has opened a public consultation on its proposal to keep young children off social media, gathering comments on the so-called Kids Act until 26 November, while Meta has used the same weeks to publish new AI systems aimed at child sexual exploitation and to extend its school partnership programme on Instagram to nearly 17,000 American schools.

The consultation, announced by the European Commission, invites children themselves, along with parents, guardians, teachers, educators and the platforms covered by the proposal, to submit observations. The feedback will feed into the material Brussels presents to the European Parliament and the Council during the legislative negotiations.

What the Kids Act would change

The Kids Act is a proposal adopted by the Commission in September, not a law in force. Ursula von der Leyen announced its key points in her State of the Union speech in Strasbourg, before the official presentation on 17 September 2026.

At its centre is age. No personal social media account would be allowed for anyone under 13. Between 13 and 15, only supervised "mini accounts" would be permitted, with reduced functions, limited usage time and direct parental control. Reported details of the supervision regime include a daily usage cap of one hour for that age group. From 15 to 18, the mandatory controlled setup falls away, but platforms would remain bound by binding "safe design" standards.

The proposal also reverses the burden of proof. Instead of users or authorities having to demonstrate after harm that a platform's design caused it, the companies themselves would have to show that their services are safe by default and appropriate for the age of their users.

The design provisions target the mechanics that keep young users on the screen. The proposal puts under scrutiny continuous scrolling, autoplay, persuasive notifications, streaks and recommendation systems that can encourage compulsive use. The Commission also wants reliable age verification that goes beyond a self-declared date of birth, and stronger parental control tools.

The rules would reach beyond familiar social networks such as Instagram and TikTok. Online video games and AI chatbots are expected to fall within scope as well, with the stated aim for chatbots of preventing functions that could foster inappropriate emotional relationships between minors and virtual assistants.

Enforcement is expected to follow the model already used under the Digital Services Act and the AI Act, with the Commission overseeing the highest-risk large platforms and national authorities supervising smaller operators. Companies found responsible for harm to minors could face fines of up to 6 per cent of annual worldwide turnover, and, in the most serious scenario, restrictions on access to the single market.

Meta's safety announcements

As the consultation opened, Meta published an update on 7 October describing new Large Language Model systems designed to detect what it calls signposting: advertisements that appear harmless but use indirect signals to steer users towards child sexual abuse material or illegal activity hosted off the platforms. The measures include deeper analysis of link destinations, new AI scans of advertising content, systems to prevent the recreation of removed accounts, and an AI red teaming agent tasked with finding weaknesses in the company's defences before they can be exploited.

Meta also stated that between January and June 2026 it acted on 33.2 million items of content related to child sexual exploitation on Facebook and Instagram, with more than 97 per cent detected before any user report. Those figures come from the company itself and measure content acted upon, not victims or independent incidents.

The second Meta announcement, on 29 September, extended Instagram's School Partnership Program in the United States to nearly 17,000 schools. Verified schools gain Channels, School Stories, Notes, reports with priority review and educational materials. High school students verified through UNiDAYS get school directories, profile banners, clubs, teams and information spaces tied to their school community. Administrators and educators can act on school-related accounts and content, while the Family Center tells parents about their child's participation.

The wider picture of claims and proceedings

The Kids Act arrives while platforms already face legal and regulatory proceedings on several fronts. The Commission has issued preliminary findings, not final decisions, in cases under the Digital Services Act. On 6 February 2026 it said it had found, for now, that TikTok breaches the Act through addictive design, naming infinite scroll, autoplay, push alerts and personalised feeds. On 10 July 2026 it issued preliminary findings that Meta breaches the same law through addictive design, saying it expects autoplay and infinite scroll to be off by default, screen time breaks, and a feed less driven by engagement. A fine in either case would be capped at 6 per cent of worldwide yearly turnover.

Whether a group claim could be brought in England and Wales is a separate matter governed by UK procedural rules, including the possibility of a group litigation order. See who can claim.

What happens next

The consultation on the Kids Act closes on 26 November 2026. After that, the proposal enters negotiations between the European Parliament and the Council, and the legislative process could take years before the text becomes law applicable across all 27 member states. In the meantime, the Commission's preliminary findings against Meta and TikTok under the Digital Services Act remain open, and the US trials and settlements continue.

Anyone affected by the issues in this article can contact Samaritans free on 116 123 or at samaritans.org.