Proposed: 17.09.2026     |     Applies to: Services reaching EU Minors     |     Builds on: DSA and AI Act

The EU KIDS Act: four pillars, what changes for platforms, and where a Trust and Safety team should start preparing.

On 17 September 2026, the European Commission proposed the EU KIDS Act, a law meant to make the internet safer for children. It has not passed yet. The proposal now goes to the European Parliament and Council, a process that typically runs a year or more before anything becomes binding. Platforms that wait for the final vote will be starting late.

The proposal follows a report from the Special Panel on Child Safety Online, a group of more than 60 experts, including child psychiatrists, neurologists, and youth representatives, convened by von der Leyen and delivered in July 2026. The Commission also points to its own Eurobarometer survey, where 92 percent of Europeans named stronger protection for children online a top policy priority.

The burden of proof moves to platforms

Under current EU rules, regulators generally have to prove a platform caused harm before they can act. The KIDS Act flips that. Platforms will have to prove their services are safe by design, before problems appear.

That is a real shift in who carries the risk. It applies to social media, video sharing, online video games, AI companions, and chatbots, wherever the audience includes anyone under 18.

Platforms now have to prove their services are safe by design, before problems appear.

The core shift at the center of the KIDS Act

Gaming platforms are named directly

The KIDS Act names online video games specifically, alongside social media and AI companions. That is different from earlier EU rules, which focused mostly on social platforms.

If a product includes voice chat, an AI companion, guild or clan features, or any feed that ranks content for players, the safety by design pillar applies as directly as it applies to a social app.

Four pillars carry the law

Social media delay

A gradual, age-based ramp into social media. What a child can do depends on which band they fall into.

0 to under 3 3 to under 13 13 to under 15 15 to under 18
No screens. No social media. No personal devices. Can use child-friendly video-sharing through a guardian’s account. Max 1 hour a day. Guardian sets up a mini account on the guardian’s own account. No unwanted contact. Max 1 hour a day. Can set up their own social media accounts.

Safety by design

Obligations for all online services offering social media, video sharing, online video games, AI companions, and chatbots to anyone under 18.

  • No addictive features, no infinite scroll without stopping points, no reward tricks, and no push notifications during sleeping hours.
  • No profiling-based feeds that pull minors into rabbit holes of harmful content, and no unsolicited contact from strangers.
  • AI companions and chatbots off by default, and they cannot simulate relationships in ways that create emotional dependency.
  • Profiles private by default, with geolocation, camera, and microphone access turned off.
  • Easy ways to block and mute, time-management tools, and recommender systems minors can control, tune, and reset.

Privacy preserving age assurance

Online services and app stores must use age assurance tools. The Commission points to its own EU age verification app, which does not retain identity documents or biometric data.

  • Social media and video-sharing providers must verify age when a user opens a new account.
  • For existing accounts, providers estimate age using reasonable proxies, such as the account creation date or credit card details.

Effective enforcement

The burden of proof sits with providers. Very large online platforms have to show their services are safe by design.

  • Very large platforms submit a compliance plan and pass an independent audit. The Commission can request corrective measures if the auditor finds shortcomings.
  • Enforcement builds on the structures already in place under the Digital Services Act and the AI Act.
  • Investigations into non-compliance are meant to conclude within 90 days.

Where a moderation platform fits

Pillars 2 and 4 assume a platform can catch harm early, act on reports quickly, explain its decisions, and show its work when a regulator asks. That is standing infrastructure, in place before anyone comes knocking. Here is where Amanda’s existing capabilities line up.

Amanda helps you

Behavior pattern detection

Surfaces accounts showing grooming or predatory patterns early, before harm happens

Trusted flagger prioritization

Reports from designated trusted flaggers surface ahead of the general queue.

Appeal and explanation flow

Minors and guardians can understand and contest a moderation decision.

Transparency reporting

Generates the transparency reports you’re already required to publish under the DSA.

Risk assessment template

A starting structure for risk documentation, whatever your regulator asks for.

Two important questions:

Will it apply outside the EU

It applies to any platform offering services to users in the EU, regardless of where the company is based, the same territorial reach as the Digital Services Act.

When does it take effect

Not yet. It is a proposal submitted to Parliament and Council on 17 September 2026. It needs their approval first, which typically takes a year or more.

Get started now

The KIDS Act still has to clear Parliament and Council. That process could change details, but it will not change the direction. Platforms serving minors in the EU are heading toward safety by design as a legal requirement.

Mapping a current moderation and reporting setup against these four pillars now turns the wait for Parliament and Council into preparation time.

This piece covers what’s changing. The DSA guide covers the reporting, appeals, and transparency work you need today.