Artificial intelligence is quickly becoming part of everyday life. Children and teenagers are using AI to answer homework questions, ask for advice, talk through personal problems, and sometimes simply have someone, or something, to talk to.
California lawmakers are now confronting a difficult question: What happens when an AI chatbot stops functioning like a tool and begins functioning like a companion?
California has enacted some of the country’s strongest protections governing AI companion chatbots, including new requirements specifically designed to protect children and teenagers.
The changes follow growing concerns about minors developing emotionally intense relationships with AI systems and reports of chatbots engaging in conversations involving suicide, self-harm, sexual content, and other potentially dangerous subjects.
Here is what California parents should know.
What Is an AI Companion Chatbot?
Not every chatbot falls into the same legal category.
California law focuses in part on “companion chatbots,” meaning AI systems capable of producing human-like responses and interacting with users in ways that can satisfy social needs.
These systems can be different from a traditional customer-service bot that simply tells you when a store closes or helps you track an order.
A companion chatbot can carry on extended conversations. It may remember information, respond emotionally, simulate friendship, provide advice, or otherwise create an experience that feels more personal.
That ability is precisely what has attracted the attention of California lawmakers.
Why Did California Regulate AI Chatbots?
The issue became particularly urgent following reports involving teenagers who developed extensive relationships with AI chatbots.
One of the cases at the center of California’s latest legislation involves Adam Raine, a California teenager who died by suicide in 2025.
His parents subsequently sued OpenAI. Their lawsuit alleges that interactions with ChatGPT contributed to their son’s death, including allegations concerning conversations about suicide and methods of self-harm.
Those allegations are part of pending litigation and should not be confused with a final judicial determination that OpenAI caused his death.
But the case helped bring national attention to a broader question: What responsibility should an AI company have when its system recognizes that a child may be considering suicide or seriously harming himself or herself?
California lawmakers have now answered part of that question through legislation.
Why Does It Matter That AI Must Say It Isn't Human?
At first glance, requiring a chatbot to tell someone that it is artificial intelligence may sound obvious.
But the concern becomes more significant when a person interacts with the same AI repeatedly.
Modern AI can respond conversationally, remember context, use emotional language, and adapt its responses based on what a person says.
For a child or teenager, those characteristics may make the interaction feel increasingly personal.
California’s disclosure requirements are intended to interrupt that effect by periodically reminding young users that there is no human being on the other side of the conversation.
What Is Adam's Law?
California strengthened its protections again in September 2026.
Governor Newsom signed a package of legislation addressing children’s interactions with technology, including Senate Bill 1119.
The legislation has been referred to as “Adam’s Law” in memory of Adam Raine.
The law imposes additional child-safety requirements on companies operating companion chatbots.
Among its significant features are requirements involving child-safety risk assessments, independent safety audits, parental controls, and procedures for situations in which a minor expresses an intention to harm himself or herself.
The legislation represents a significant development because California is moving beyond simply requiring warnings.
The state is increasingly placing affirmative safety responsibilities on companies that design and operate AI systems used by children.
What Happens If a Child Tells an AI Chatbot About Suicide?
This is one of the most important parts of California’s approach.
AI companies covered by the laws cannot simply treat statements about suicide or self-harm as ordinary conversation.
California’s chatbot regulations require protocols for addressing suicidal ideation and self-harm.
Under the strengthened protections enacted in 2026, covered chatbot operators also face requirements involving parental notification when a child’s interactions indicate certain threats of self-harm.
That represents an important shift in the law.
The question is no longer merely whether an AI chatbot should respond appropriately when a child discusses suicide.
California is also regulating what the company operating that chatbot must do after its system identifies a potentially dangerous situation.
California Is Requiring More Accountability From AI Companies
Another important development is the increased emphasis on auditing and risk assessment.
Under the newer protections, covered companies must conduct assessments designed to identify risks their AI systems may present to children.
Independent child-safety audits are also part of the regulatory framework.
This matters because a company’s public promise that its product is safe is different from requiring the company to systematically evaluate risks and subject its practices to outside scrutiny.
California is increasingly treating child safety as something that must be incorporated into the design and operation of AI products rather than addressed only after something goes wrong.
Can AI Companies Be Held Legally Responsible?
Potentially.
California’s developing AI laws create regulatory requirements that companies must follow, and some provisions create potential avenues for legal liability when covered companies fail to comply.
But liability in an individual case is not automatic.
A lawsuit involving an injury or death can raise complicated questions about duty, causation, foreseeability, statutory requirements, product design, warnings, and the conduct of multiple parties.
The existence of an AI conversation alone would not necessarily establish that an AI company is legally responsible for a subsequent injury.
At the same time, the creation of specific statutory safety obligations could become important in future litigation involving AI companies.
The Law Does Not Apply to Every AI Interaction in Exactly the Same Way
It is important not to describe California’s legislation as simply regulating “all AI.”
The laws contain definitions determining which systems and companies are covered.
For example, California has distinguished companion chatbots from certain systems designed for narrower purposes, such as some customer-service applications.
Different requirements can also apply depending on whether the user is a minor.
As AI becomes incorporated into search engines, phones, social media platforms, toys, educational products, and other technology, determining which California law applies may itself become a significant legal question.
What Should California Parents Know?
Parents should understand that AI chatbots are not necessarily the same thing as search engines.
A teenager may begin by asking an AI system a simple question and eventually begin discussing relationships, depression, loneliness, sexuality, family problems, or other deeply personal subjects.
The technology may respond in a conversational manner that feels very different from reading information on a traditional website.
California’s new laws recognize that distinction.
Parents should also understand that the law is developing quickly. Protections that did not exist several years ago are now becoming mandatory, and California enacted additional requirements in both 2025 and 2026.
AI Safety Is Becoming a Legal Issue, Not Just a Technology Issue
For years, discussions about artificial intelligence largely focused on what the technology could do.
The legal conversation is increasingly focused on what companies should be required to prevent.
Should an AI chatbot recognize when a child is in danger?
Should parents be notified?
Should companies be required to test their systems for foreseeable risks to minors?
Should children be reminded that the seemingly sympathetic person responding to them is actually software?
And when a company knows its technology is being used by children, what legal responsibilities should follow?
California is beginning to answer those questions through legislation.
Frequently Asked Questions About California's AI Chatbot Laws
Does California regulate AI chatbots?
Yes. California has enacted laws specifically regulating certain AI and companion chatbot systems, including protections relating to minors, suicide and self-harm, disclosures, sexually explicit content, safety assessments, and other risks.
Does an AI chatbot have to tell a child that it isn't a real person?
California’s chatbot protections include disclosure requirements intended to make minors aware that they are interacting with artificial intelligence rather than another human being.
Are there special rules when a minor discusses suicide with an AI chatbot?
Yes. California law requires covered chatbot companies to maintain protocols addressing suicidal ideation and self-harm. Newer protections impose additional responsibilities involving children and potentially dangerous interactions.
Can parents control their child's AI chatbot use?
California’s newer child-safety legislation includes parental-control requirements for covered companion chatbot services.
Can an AI company be sued if a chatbot harms someone?
Potentially, depending on the circumstances. California law can provide legal consequences for violations of applicable requirements, but whether a company is liable for a particular injury depends on the facts and the applicable causes of action.
Are ChatGPT and every other AI chatbot governed by exactly the same rules?
Not necessarily. California statutes define the systems and operators subject to particular requirements, and exceptions or different requirements may apply depending on the type of AI service and how it is being used.
The Bottom Line
California is moving toward a legal framework in which companies cannot simply release increasingly human-like AI companions and leave children to navigate the risks themselves.
The state first imposed significant companion-chatbot safeguards through legislation signed in 2025. In September 2026, Governor Newsom signed additional legislation strengthening protections for children, including Adam’s Law.
The emerging principle is significant: the more technology is designed to communicate like a person, build relationships with users, and respond to deeply personal information, the more lawmakers are asking whether the companies behind that technology should have corresponding responsibilities to protect vulnerable users.
For parents, attorneys, technology companies, and anyone using AI, these laws are worth watching.
Artificial intelligence may be evolving quickly, but California law is beginning to catch up.
Disclaimer
This blog is for informational purposes only and does not constitute legal advice. For specific legal guidance on your case, please contact The Mines Law Firm for representation.