Skip to main content

EU KIDS Act: Proposed EU Rules for Online Platforms, Age Verification, and AI Chatbots

Read in German.

On Sept. 17, 2026, the European Commission proposed the EU KIDS Act, a new regulation intended to strengthen protections for minors using digital services and AI systems. The EU KIDS Act is currently a legislative proposal and not yet applicable law. If adopted, the regulation would introduce EU-wide requirements for age verification, child-safe product design, contact protections, and AI companions and conversational chatbots, including for providers established outside the EU. The proposal is particularly relevant for social media platforms, video-sharing services, online games, app stores, operating system providers, and providers of companion and general-purpose conversational AI systems.

Key Takeaways for Digital Service Providers

  • Independent accounts on covered social networks and video-sharing platforms with “risky” functions would generally be prohibited for users under 15.
  • Certified, privacy-preserving age verification would be required for key account restrictions.
  • Covered services would face child-safe design, contact protection, and default-setting requirements.
  • AI companions and conversational chatbots would be subject to specific design, testing, monitoring, and parental-control duties.
  • The proposal would apply to certain providers outside the EU and might expose AI companion and chatbot providers to fines of up to 6% of worldwide annual turnover.

  • Scope of the EU KIDS Act: Covered Services and Providers

    The EU KIDS Act is a proposed regulation based on Article 114 of the Treaty on the Functioning of the European Union (TFEU).

    The proposal would apply to a broad range of digital services, including those offered by providers established in third countries. The services within scope include:

    • social networks;
    • video-sharing platforms;
    • online games;
    • app stores;
    • operating systems, with respect to the transmission of age signals; and
    • AI companions and general conversational chatbots.

    Not-for-profit online encyclopedias and educational or scientific repositories, qualifying educational and research services, certain public-authority services, and open-source software development and sharing platforms would be excluded, subject to the conditions set out in the proposal. Conversational AI systems limited to specialized functions, such as certain customer-service and technical-support applications, would not qualify as general conversational chatbots. Micro and small enterprises are not generally exempt, although certain monitoring duties for AI chatbots would not apply to them.

    Key EU KIDS Act Requirements for Digital Service Providers

    Age Limits and Rules for Existing User Accounts

    For covered social networks and video-sharing platforms with risky functions, independent accounts would be prohibited below the age of 15. For 13- and 14-year-olds, guardians may set up accounts with limited functions, subject to a maximum of one hour daily, parental control, and contact approval. On these services, no accounts would be permitted for children under 13. Child-specific video-sharing services may, by way of exception, allow accompanied access through a guardian’s account, limited to one hour daily and not for children under the age of three.

    Within six months of the date of application, providers would need to clarify the age of existing account holders and block those held by users under 15 or whose age cannot be determined, unless the provider could establish the minimum age with a high degree of confidence.

    Certified Age Verification and Parental Responsibility

    Self-declaration would not suffice. For the account restrictions, only certified, privacy-preserving EU age-verification solutions (zero-knowledge principle) would be permitted. For the safety requirements, other methods would be allowed if they met equivalent standards. Member States would be required to provide a free verification solution and a means to demonstrate parental responsibility.

    Child-Safe Design Requirements for Online Platforms

    The proposal would require social networks and video-sharing platforms to be designed with child safety as a default.

    Key obligations include:

    • a prohibition on design features that encourage compulsive or excessive use by minors, including autoplay and uninterrupted content consumption without effective breaks, certain re-engagement notifications, and reward mechanisms encouraging repeated engagement;
    • time limits that protect school time and core sleeping hours;
    • recommender systems that prioritize safety and mental health, with a mandatory non-profiling option and engagement-based signals disabled by default;
    • safe default settings, for example location, camera, microphone, and push notifications switched off;
    • a prohibition on exposing minors to variable reward systems on social networks and video-sharing platforms; and
    • child-friendly reporting channels and tools for guardians.

    Online games and AI chatbots would be subject to tailored subsets of these requirements. App stores would have to apply age ratings and block minors’ access to apps that are not age appropriate.

    Contact and Privacy Protections for Minors

    Direct contact with minors would be possible only after prior approval, addition to groups would require explicit consent, and minors would have to be able to block other users without disclosing their identity to the blocked user. Minors would not be included in contact recommendations, and their profiles would not be visible to strangers by default.

    Requirements for AI Companions and Conversational Chatbots

    Providers would need to avoid design features capable of creating emotional dependency through simulated interpersonal relationships. The proposal envisions pre-market risk testing and post-market monitoring (the latter not for micro and small enterprises). The use of information or analysis derived from minors’ prior interactions in subsequent conversations would be disabled by default, except where necessary to protect the minor’s safety, and access for users under 13 would be possible only through parental control.

    Enforcement, Fines, and Supervisory Fees Under the EU KIDS Act

    The proposal builds on the existing enforcement frameworks under the Digital Services Act (DSA) and the EU AI Act, with supervisory responsibilities divided between the European Commission and national competent authorities. Data protection authorities would separately supervise compliance with the data protection requirements for age assurance and may impose fines under the GDPR. Very large social networks and video-sharing platforms would be required to have a compliance plan assessed by independent auditors.

    The financial consequences would include:

    • Fines: for infringements of the regulation, up to 6% of worldwide annual turnover for AI companions and chatbots; for other services, the DSA sanction regime would apply.
    • Supervisory fee: for providers under Commission supervision, up to 0.03% of worldwide annual net income, charged to cover supervision costs (not a penalty).

    Minors and guardians would have a right to complain and may mandate organizations to act on their behalf.

    Legislative Timeline and Next Steps

    The EU KIDS Act remains a proposal and may change during the legislative process. As currently drafted, the regulation would generally apply six months after its effective date, with certain exceptions. In particular, the provisions on compliance plans for very large platforms would apply from the effective date, while certain national support measures and expedited enforcement procedures would apply after 12 months. The Commission might also amend certain requirements by delegated acts. If adopted as proposed, it would reshape how in-scope services onboard minors, design their products, and govern AI chatbots. The Commission’s review is currently anticipated by Aug. 31, 2030.

    What Digital Service Providers Should Consider Now

    Companies should consider assessing whether their services may fall within the proposed scope; reviewing age-assurance, age-verification, and onboarding processes; identifying potentially addictive or engagement-based design features; examining recommender systems and default settings; and reviewing safeguards governing minors’ interactions with AI companions and conversational chatbots. Any preparations should account for the fact that the proposal may change during the legislative process.



     Special thanks to Paralegal Felix Winter ˘ for contributing to this GT Alert.

    ˘ Not admitted to the practice of law.