EU Kids Act could require age checks for social media and AI chatbots

Online platforms including social media, video-sharing services, certain online games and AI chatbots could be required to verify users’ ages when new accounts are created under a proposed European Union law aimed at strengthening child safety online.

The European Commission is due to present the EU Kids Act on 17 September 2026, with the proposal expected to establish an EU-wide framework for age-based access to a range of digital services. Commission officials confirmed on 14 September that the draft was being discussed by commissioners ahead of its planned presentation, but declined to disclose details before the formal announcement.

Details of the proposed system have emerged from internal Commission documents obtained by Euractiv. Because the documents concern a proposal that had not yet been formally presented as of 16 September, the provisions could still change.

Proposed minimum age of 15 for autonomous accounts

The draft would establish 15 as the minimum age for a minor to open an autonomous account on covered social media and other higher-risk services, according to the documents reported by Euractiv.

The proposed system would use different levels of access according to age.

Children under three would have no access to social media or services classified as risky to minors. Children aged three to under 13 could use child-friendly services through accounts fully controlled by their parents.

For those aged 13 to under 15, the draft proposes an introductory account for social media and video-sharing platforms, with parents responsible for setting up and controlling access. From the age of 15, minors could open autonomous accounts where the online environment meets the proposed "safe by design" requirements.

The measures would extend beyond conventional social networks. The draft identifies social media and video-sharing platforms, risky online games, and AI chatbots and companions as services that could fall within the wider category covered by the legislation.

Services operated for educational purposes, public authorities and certain industrial or office-related AI applications would be outside the proposed scope, according to the documents reported by Euractiv.

Age verification would apply when accounts are opened

The draft would require providers of social media services and video-sharing platforms to carry out age verification when users create new accounts.

Providers could use the EU's age-verification system or another solution prepared by public authorities.

The Commission has already developed an EU-wide technical approach to age verification. Its feature-ready solution became available in April 2026, and the Commission says it is designed to allow users to prove that they meet an age requirement without unnecessarily revealing additional personal information.

A Commission recommendation adopted in April calls on EU member states to make an EU age-verification solution available by the end of 2026. The recommended system is intended to use privacy-preserving technologies and, where possible, provide services with confirmation that a person meets an age threshold rather than additional identity information.

The existing Digital Services Act already requires online platforms accessible to minors to implement appropriate and proportionate measures to ensure a high level of privacy, safety and security for children. The Commission's 2025 guidelines also recommend measures including safer default account settings, restrictions on unwanted contact and changes to recommender systems.

‘Safe by design’ rules would target addictive features

The proposed legislation would also introduce requirements for services covered by the new rules to be designed with children's safety in mind.

The draft criteria reported by Euractiv include restrictions on features such as infinite scrolling, greater controls over recommender systems and safer default account settings.

The Commission has already taken regulatory action concerning the design of major online platforms. In July 2026, it said its preliminary findings indicated that Facebook and Instagram had features potentially breaching the Digital Services Act, including infinite scrolling, autoplay, push notifications and highly personalised recommender systems.

The proposed Kids Act would therefore build on an existing EU regulatory framework rather than create the bloc's first rules governing children's online safety.

AI chatbots and companions included in wider approach

AI systems are a significant addition to the proposed framework.

The Commission's existing guidance on protecting minors online already recognises that the growing use of AI chatbots and companions can create additional privacy, safety and security risks for children. Its 2025 guidelines specifically recommend safeguards around AI chatbots integrated into online platforms.

The European Parliament has also previously called for a harmonised EU digital age limit covering social media, video-sharing platforms and AI companions, alongside stronger privacy-preserving age-assurance systems.

Under the draft Kids Act, AI chatbots and companions classified as risky to minors would therefore be addressed alongside social media, video-sharing platforms and certain online games.

Commission preparing enforcement system

The proposed legislation would create additional enforcement costs for covered companies.

According to the internal documents reported by Euractiv, services designated under the new framework would be subject to a supervisory fee to help fund the European Commission's enforcement activities.

The proposal would add another layer to an EU digital regulatory system that already includes the Digital Services Act, which gives the Commission direct supervisory responsibilities over the largest online platforms and search engines.

The exact structure of the proposed fee and enforcement arrangements will depend on the final legislation presented by the Commission.

Proposal follows months of EU work on child safety

The Kids Act follows a year of EU work on children's online safety.

In 2025, Commission President Ursula von der Leyen announced the creation of a special expert panel to examine children's online experiences and possible European approaches to age restrictions.

The panel met between March and June 2026 and submitted its final report in July. Its work covered social media, gaming, messaging applications and AI tools, as well as the responsibilities of technology companies and possible EU-wide age limits.

A Commission survey published in July found that 71% of respondents identified cyberbullying and harassment as major concerns for children online, while 70% cited online grooming and sexual exploitation. Nearly two-thirds supported EU rules restricting children's access to social media according to age.

A separate Commission survey published in June found that young people in the EU spent an average of 4.5 hours online on school days and more than six hours at weekends.

Formal proposal due on 17 September

The Commission is scheduled to present the EU Kids Act on Thursday, 17 September, following discussions among commissioners and President von der Leyen's State of the Union address on 16 September.

Commission spokesman Olof Gill confirmed on 14 September that the Commission would provide details at the later presentation and declined to pre-empt the contents of the proposal.

If adopted by the Commission, the legislation would then enter the EU's ordinary legislative process, requiring negotiations between the European Parliament and the Council of the EU before it could become binding law.

Until the formal proposal is published, the age thresholds, categories of services, enforcement mechanisms and other provisions described in leaked or internal documents remain subject to change.