AI Chatbot Rules for Kids: California, UK and Australia -

Quick answer: On 10 September 2026, California signed “Adam’s Law,” the strictest US rules yet for AI chatbots used by children. Operators will have to check ages, limit sessions and send crisis resources, with most duties starting on 1 July 2027. The UK and Australia already have their own chatbot rules, and a federal US bill is still waiting for a full Senate vote as of the latest reports we found.

Governments on three continents are now treating AI chatbots, especially “companion” apps that act like a friend, as a child-safety issue. The rules differ a lot by country, so what your teen can use depends partly on where you live.

Below is what each place has done, what’s confirmed, and what parents can do today. For more on how these tools work, see our AI news and guides.

What does California’s Adam’s Law require?

Governor Gavin Newsom signed 13 child online-safety bills on 10 September 2026, according to CalMatters. Adam’s Law (SB 1119), written by Senator Steve Padilla, is the chatbot one. It’s named after Adam Raine, a teenager whose family says a chatbot gave harmful responses.

A legal analysis by ZwillGen says the law takes effect on 1 January 2027, with most operating duties starting on 1 July 2027. Companies that offer companion chatbots must either verify each user’s age or apply child protections to everyone. The protections include:

Risk assessments before release and published child-safety policies.
Crisis protocols that point users to 988 or an equivalent service.
Default settings with memory switched off, one-hour sessions and a two-hour daily cap.
Periodic reminders that you’re talking to AI.
Limits on behavioral advertising and data sales.

The law also bars chatbots from claiming to be sentient, expressing romantic interest or offering excessive flattery. Civil penalties reach $5,000 per affected child for negligent violations and $15,000 for intentional ones, and families can sue.

The same package includes a four-year moratorium on AI chatbot toys for children under 16, per CalMatters. Courthouse News confirms the signing date and the 13-bill count.

What rules already apply in other US states?

California’s earlier law, SB 243, took effect on 1 January 2026. Jones Walker summarizes it: operators must tell users they’re talking to AI, remind minors to take a break at least every three hours, block sexual content for minors and refer people who express suicidal thoughts to crisis services. Damages are $1,000 per violation or actual damages, whichever is greater.

At the federal level, the GUARD Act (S. 3062) would bar minors from AI companions and require age verification. The Senate Judiciary Committee approved it unanimously on 30 April 2026, per Covington’s Global Policy Watch. We couldn’t find a report that it has passed the full Senate, so treat it as pending.

What is the UK doing about AI chatbots?

The UK is extending its Online Safety Act. Stephenson Harwood reports that the Crime and Policing Act 2026 and the Children’s Wellbeing and Schools Act 2026 both received Royal Assent at the end of April 2026. The first adds a power to bring AI-generated illegal content within the Online Safety Act and criminalizes nudification tools. The second lets ministers set screen-time limits, curfews or age restrictions for children’s use of services.

Those powers still need regulations. According to the same firm, the government has up to 12 months after a progress report to make them, so don’t expect UK teen limits to appear overnight. The government’s initial announcement in February 2026 followed controversy over deepfake images from the Grok chatbot, as GlobalPost reported.

What about Australia?

Australia’s eSafety Commissioner says its age-restricted material codes already apply to AI companion chatbots. They must stop their service generating sexually explicit content for children or use appropriate age assurance, and provide crisis and mental-health information. An eSafety report found that the four services it examined relied on self-declared ages. Breaching a compliance direction can bring civil penalties of up to A$49.5 million.

After eSafety’s notices, Character.AI removed open-ended chat for under-18 users and Chub AI withdrew from Australia, the regulator says.

What are the limits of these laws?

Laws set a floor, and they take time to bite. Adam’s Law doesn’t start its main duties until mid-2027, UK ministers still have to write the regulations, and the US federal bill hasn’t passed. In the meantime, the protections your child gets depend on each company’s own settings.

Age checks are also imperfect. Australia’s eSafety Commissioner found that the services it reviewed mostly relied on users stating their own age, which is easy to get around. Expect rules to push companies toward stronger checks, and expect debate about privacy when they do.

Finally, these laws focus on companion apps and chatbots aimed at, or open to, children. They don’t replace a conversation at home about what chatbots can and can’t do. A chatbot isn’t a therapist, a doctor or a friend, however friendly it sounds.

What can parents do now?

Find out which chatbots your child uses. Companion apps and general assistants behave differently. Ask, don’t just check.
Use accounts with the correct age. Protections tied to age only work when the age is right.
Turn off memory and chat history where the app lets you, and review privacy settings together.
Agree on a rule for hard topics. Make it clear your child can come to you, not just to a chatbot, about stress or health.
Know the crisis numbers. In the US and Canada you can call or text 988. In the UK, Samaritans is on 116 123. In Australia, Lifeline is 13 11 14.

Adults can benefit from the same discipline. Our guide to https://insightorys.com/2026/10/07/use-ai-safely-at-work/ covers what to keep out of any chatbot, and our explainer on https://insightorys.com/2026/10/07/eu-ai-act-labels-2026/ covers the EU’s new labelling rules.

Frequently asked questions

Is Adam’s Law in force yet?

Not yet. It was signed on 10 September 2026 and takes effect on 1 January 2027, with most operational duties starting on 1 July 2027, according to ZwillGen’s analysis.

Does Adam’s Law apply outside California?

Legally, it covers operators offering companion chatbots in California. Companies may roll out changes more widely, but none has confirmed that to us.

Is there a federal US law on AI chatbots and kids?

Not as of the latest reports we found. The GUARD Act cleared a Senate committee on 30 April 2026 but still needs the full Senate and the House.

Does Canada have a similar rule?

We haven’t confirmed a chatbot-specific federal law in Canada, so we’ve left it out rather than guess.

AI Chatbot Rules for Kids: California, UK and Australia