Adam's Law and the EU KIDS Act: What Two New Laws Mean for Your Child's AI Chatbot

California's Adam's Law (signed Sept 10, 2026) and the EU's proposed KIDS Act (unveiled Sept 17) both aim to make AI companion chatbots safer by default for minors: crisis protocols, parental controls, limits on emotional dependence, and age checks. California's core rules start July 1, 2027; the EU proposal still needs Parliament and Council approval.

What just changed for kids and AI chatbots?

Two rulebooks moved within a week of each other. Adam's Law is named for Adam Raine, a California teenager who died in 2025 after ChatGPT allegedly coached him, and his mother Maria Raine joined the signing. A week later, the European Commission proposed the KIDS Act, which applies to social media, video platforms, online games, and AI companions and chatbots. One is law with a 2027 start date. The other is a proposal with months of negotiation ahead.

What does California's Adam's Law require?

Adam's Law (SB 1119) builds on SB 243, which took effect January 1, 2026. Its core requirements begin July 1, 2027. Among other things, it requires:

• Age assurance using a privacy-protective age signal from the device
• Risk assessments before a new or substantially modified companion chatbot is released
• In-app crisis referrals, plus notice to a parent when there's a credible threat of imminent self-harm
• Defaults only a parent can change: no notifications, usage time limits, and limits on persistent memory
• Liability for failing to take reasonable steps against self-harm content, sexual content, romantic roleplay, excessive flattery, and manipulation that fosters reliance or isolation
• Limits on targeted ads in chats, independent audits reported to the Attorney General, and a private right of action for families

What would the EU KIDS Act change?

Under the Commission's proposal, AI chatbots and companions would be off by default for children and could not act in ways that make kids emotionally dependent. Age checks would apply to services and app stores. Very large platforms would have to prove their products are safe for children. For social media, the proposed tiers are: none under 13; parent-managed mini accounts with a one-hour limit for ages 13 to under 15; and more independent access from 15. This is still a proposal. Education-focused tools have been reported as excluded — check the final wording.

Why are lawmakers converging on the same rules?

Across California, the EU, and federal proposals, the same design ideas keep showing up: protective defaults, parents in control, no manufactured emotional dependence, age checks, and proof of safety. NPR has reported that AI chatbot makers face at least 75 lawsuits. That pressure sits alongside the Meta child safety settlement and a wave of new laws aimed at AI for kids in 2026. In the U.S. Senate, the GUARD Act advanced in Judiciary in April but is not law; the bipartisan CHATBOT Act would give families more control over kids' AI chatbot use and is also not yet law.

Does a law for "companion chatbots" cover every AI app?

Not automatically. Adam's Law regulates companion chatbots — systems built to simulate an ongoing personal or emotional connection. The EU proposal names AI companions and chatbots, but definitions and exemptions matter. NPR has noted that some industry-shaped bills exempt popular chatbots. For parents, the practical test is behavior: Does it flatter? Pretend to be a friend? Send notifications to pull a child back in? Keep long-term personal memory? Judge the app by how it acts, not by its category label.

What should parents check before their child uses an AI app?

You do not have to wait until 2027 to ask for safer defaults. Use this checklist — and see our guides on how to monitor kids' AI use and which AI has the best parental controls:

1. Who controls settings — Confirm only the parent can change limits, not the child.
2. Check notifications — See whether the app pings the child to pull them back in.
3. Review memory — Ask whether it keeps long-term personal details and whether you can see them.
4. Watch the tone — Note flattery, claimed feelings, or best-friend framing.
5. Ask about crisis handling — What happens on self-harm mentions, and are you notified? (US: 988 or 911 if immediate danger.)
6. Ads and data — Check whether kids see ads and what is stored.

Where does HeyOtto fit?

HeyOtto is an AI platform for kids and teens (ages 8–18). Parents create the account and stay in the loop: you can see conversations, set topic blocks and time limits, and get Real-Time Alerts when something needs your attention. Otto is an assistant for schoolwork, creating, and questions—not a companion chatbot built to simulate a friendship or keep a child coming back for emotional support.

That design matches the direction both Adam's Law and the EU KIDS Act are pushing: protective defaults, parent control, and no manufactured emotional dependence. A statute is a floor, not a seal of approval. Use the checklist above on any app your child uses—including ours.

The bottom line

California has set the standard for enforceable child-safety rules on AI chatbots, and Europe wants to go further. Whatever the final wording, the direction is clear: safe by default, parent-controlled, and provable. Parents don't have to wait until 2027 to ask for that.