On September 17, 2026, the European Commission published its proposal for the "EU KIDS Act" (Keeping Internet Digital Spaces Accountable and Trustworthy Act), a new law that would join the AI Act, the DSA, and the CRA as the EU's fourth major digital regulatory framework. Its most relevant point for the enterprise AI world: it would require AI chatbots and companions to ship off by default for underage users. Euronews, IAPP, and the European Commission itself confirmed the draft's details between September 17 and 22, 2026.
The KIDS Act would apply to any large online platform offering social media, video, online gaming, or AI companions and chatbots to users under 18. Two concrete rules are already in the draft: AI chatbots and companions must be off by default for minors (the minor user, or their legal guardian, would have to explicitly turn them on), and those chatbots cannot be designed or behave in ways that foster emotional dependency in children.
The draft also includes a blanket ban on children under 13 opening social media accounts, with the possibility of restricted "mini-accounts" for the 13-to-15 age range. These rules aren't AI-specific, but they're part of the same regulatory package and give a sense of the overall direction: the EU is treating minors' exposure to conversational AI systems with the same level of concern as exposure to social media.
The text requires affected platforms to apply "safety by design" principles in their products, submit compliance plans to the European Commission, and subject any new product or service to independent audit. The proposal still needs approval from EU governments and the European Parliament before becoming law, so the actual timeline for it taking effect remains uncertain, but the process is now formally underway.
The European regulatory pattern (AI Act, DSA, CRA, and now the KIDS Act) tends to become a de facto reference for other jurisdictions, and several global companies end up applying the strictest standard across all their markets instead of maintaining separate product versions per region. If your company builds or distributes a chatbot, conversational assistant, or AI "companion" with international reach — even if your user base today is mostly Latin American — this is the kind of regulation worth monitoring before it becomes mandatory, not after.
Carlos Montiel is an enterprise AI solutions architect. He implements LLMs, Agents, RAG, and orchestrators for companies across Guatemala and Latin America. Reach out for a consultation.
Contact Carlos Montiel