What happened
Following President von der Leyen's 16 September State of the European Union address, the European Commission formally unveiled the proposed EU KIDS Act (17 September 2026). The proposal applies to online services used by minors — social media, video-sharing platforms, online games, and AI companions/chatbots — and would require: AI chatbots/companions switched off by default for minors and barred from creating emotional dependency, age-tiered social-media access (no accounts under 13, none-personal under 15), safety-by-design obligations (no infinite scroll, no reward tricks, privacy-by-default), age-verification tools for app stores, a 90-day enforcement procedure, and fines up to 6% of worldwide annual turnover for non-compliance.
Why it matters
This is the first EU-level legislative instrument to directly regulate AI chatbots/companions as child-safety products, imposing design, default-off and anti-addiction obligations on any conversational AI likely to be used by minors in the EU. It extends AI-specific 'safe-by-design' duties well beyond the AI Act's risk-classification and carries turnover-based fines.
Action needed
Providers of AI chatbots, companions, and any conversational AI plausibly used by minors in the EU should assess exposure to default-off/emotional-dependency bans and age-verification obligations, and track the legislative process for the fast-track enforcement mechanism.