What the draft says
The European Commission will propose restricting children under 15 from social media, video-sharing platforms, AI chatbots and online games, according to a Commission document reported by Reuters and AFP on 14 September. The measures make up an EU Kids Act that Commission President Ursula von der Leyen and EU tech chief Henna Virkkunen are due to present on Thursday, 17 September.
Reuters called it the most sweeping plan yet to protect children from online dangers. For the AI industry, the notable part is the grouping: general-purpose chatbots such as ChatGPT sit in the same bracket as Instagram, TikTok and YouTube, rather than in a category of their own.
Four age tiers
The draft sorts children into four groups, according to both agencies:
- Under three: no access to the services at all.
- Three to 12: accounts fully controlled by parents, and only on child-friendly services that meet strict safety standards.
- 13 and 14: parents can open “introductory” accounts on social media and video-sharing platforms, with parental controls, limited contacts and strict time limits.
- 15 and over: children can set up their own accounts.
“It needs to limit the access of tech companies to our children, and not the other way around,” the document says, according to AFP.
What companies would have to do
Social media and video-sharing platforms would have to verify a child’s age when a new account is opened, and gaming platforms before a game is downloaded, AFP reported. Companies would have to avoid addictive design and harmful feeds, give children simple ways to report harmful content, and offer parental controls that work. They would also pay supervisory fees to fund the regulators that enforce the rules.
Neither report spells out how age verification would work for AI chatbots in particular, or whether a 13- or 14-year-old could reach a chatbot through a parent-opened account in the way the draft allows for social media. Those are the details to look for when the full text is published.
Why the chatbot grouping matters
Reuters said the proposal responds to concerns about the effect on children of platforms including Facebook, Instagram, TikTok, YouTube and ChatGPT. Treating a chatbot like a social feed means the same questions would apply to both: who the user is, how old they are, and whether the product is designed to keep them there.
For chatbot providers, that would add a child-access regime on top of the obligations they already face in Europe. For parents, it would mean one set of age rules covering the apps a teenager scrolls and the assistant they talk to.
What happens next
Von der Leyen may preview the plan in her State of the Union address on Wednesday, AFP reported. After Thursday’s presentation, the Commission will have to work the draft out with EU governments and the European Parliament over the coming months before it can become law. Until that process ends nothing in it applies, and the age tiers, verification duties and fees can all change along the way.
