Millions of Americans are turning to AI chatbots for far more than just answers to questions or everyday information; they’re seeking company, comfort, and mental health support.
While many have had positive experiences, state lawmakers are starting to take notice of the risks of using chatbots this way and how it’s affecting people of all ages. Over the last few years, there have been a slew of lawsuits about the sometimes-fatal outcomes of associated with using chatbots for emotional support and advice.
And it’s not just parents or individuals who are suing tech companies over AI harm. Earlier this year, Kentucky’s Attorney General sued a chatbot company for preying on children.
At the same time, a wave of new state laws has begun regulating how AI companion and "therapy" bots interact with people. They aim to protect Americans and encourage safe use, but the laws greatly vary depending on where you live.
Before diving into what those chatbot regulations actually require, it helps to understand who's using the technology this way and why. The sheer number of people who interact with chatbots powered by large language models (LLMs) for things beyond your typical Google search explains exactly why regulators felt it was important to act.
How Many Americans Talk to AI Chatbots?
According to a report by the Congressional Research Service, about 1 in 10 U.S. adults now use chatbots for emotional support or advice, and 1 in 25 use them for companionship.
In March 2026, Sentio University released its findings from their study, Large language models as mental health resources: Patterns of use in the United States. There were several notable results:
- Nearly half (49%) of respondents used a major AI model like ChatGPT, Claude, or Gemini for therapeutic support.
- 39% said AI chatbots are as equally helpful to human therapy, with 36% finding them more helpful.
- Accessibility (90%) and affordability (70%) were cited as primary motivations for using chatbots as therapeutic providers.
- 63% reported improved mental health.
Interestingly, the Sentio survey also found that 64% of participants have used chatbots for four or more months. This means they’re engaging people for longer periods of time compared to other digital mental health apps.
Clinicians are seeing a similar pattern of widespread use. In a 2026 survey by the American Psychological Association, 77% of psychologists said their patients reported using AI for mental health support.
Who Uses Chatbots for Companionship and Advice?
Every age group in the U.S. uses AI chatbots for emotional purposes, but some use it more than others.
Early teenagers to young adults are turning to AI for companionship more and more. One study showed that nearly 20% of 12- to 21-year-olds – around 8.2 million young people – have used chatbots for advice when feeling sad, angry, or stressed, and that number is rapidly increasing.
On the other end of the age spectrum, about one-third of adults aged 50 to 80 report feeling lonely or isolated, and AI companions are now heavily marketed as a way to change that. Interestingly, New York’s Office for the Aging launched a companion device for older adults in a state-run pilot program. According to the data they collected, participants using the device, ElliQ, reported a 95% reduction in loneliness scores.
These types of positive results are also being considered by states that are trying to regulate the technology.
But a key reason behind why lawmakers are introducing regulations and signing them into law isn’t due to some generic tech-policy debate or the results of pilot programs. Many stem from lawsuits alleging AI chatbots caused mental health harm that ultimately resulted in users committing suicide.
Are AI Companion Apps for Teenagers Dangerous?
Some of the safety data that’s driving new state AI laws shows the scale of the risks teenagers face when using companion apps for advice. OpenAI has said roughly 0.15% of ChatGPT users, which is the equivalent of up to 1.3 million people worldwide, discuss suicide on the platform every week.
For parents with teens, their concerns and fears are real, and knowing what protections exist is important. California’s SB 243 is a state AI law aimed at chatbot safety. It came into effect on January 1, 2026.
While the law helps protect everyone, it’s especially significant for children. SB 243 requires companion chatbot operators to disclose that a minor is talking to AI, remind minors to take a break at least every three hours of continuous use, and block sexually explicit content. But there’s more.
A crisis-prevention protocol was integrated into California’s chatbot safety law. If a minor talks about suicide with AI, the platform must give them contact information for crisis service providers or suicide hotlines. The law also requires platforms to look for hints of suicidal thoughts in chats based on scientific methods, not just keywords.
Lawmakers are moving on platform design more broadly, too. In New Mexico, a recent ruling against Meta included orders to change how Facebook and Instagram handles young users. Between youth social media addiction and AI companionship, companies are beginning to be held accountable for the harm their technology causes children and teens.
Adult Use of Chatbots for Therapy
AI is frequently used by adults as a form of therapy, but is it safe? The answer isn’t simple. While some believe it depends on what someone is asking a chatbot to do, the American Psychological Association says replacing it for real therapy isn’t safe or recommended.
Most platforms aren’t licensed to provide such mental health services, and few state laws address adult use of chatbots. This leaves a regulatory gap for the millions of American adults using Claude, ChatGPT, and others for emotional support.
One state is an outlier though. Tennessee’s Senate Bill 1580 directly prohibits marketing AI as a “qualified mental health professional”. It came into effect on July 1, 2026, and unlike other chatbot laws, it gives individuals the right to sue for harm when they’re misled by an AI mental health product.
Do AI Companions Help Lonely Seniors?
Whether isolated and lonely seniors using chatbots for companionship is helpful or there’s a regulatory blind spot for adult use of AI is up in the air.
Research suggests there are real benefits for older adults that create these “connections”. Studies on AI companions for seniors have linked sustained interaction to measurable drops in loneliness and depression scores.
But virtually none of the new state AI laws, which were built with youth safety in mind, address elder users, cognitive decline, dependency risk, or family oversight. For a demographic already vulnerable to scams and predatory marketing, that’s a gap worth watching as the technology becomes even further integrated into everyday life.
AI Companion Laws and Wrongful Death Lawsuits
Unfortunately, new state laws were developed following numerous cases of people allegedly harmed by AI. There were several deaths connected to the use of chatbots that resulted in high-profile lawsuits, including a wrongful death case in New York.
New York’s AI Companion Models law, which came into effect on November 5, 2025, was partly the result of a mother’s advocacy. Megan Garcia’s 14-year-old son committed suicide after developing an emotional relationship with a Character.AI chatbot.
She filed a wrongful death lawsuit against Character Technologies. That lawsuit, which settled earlier this year, and others like it that involve minors using chatbots during times of crises, continue to drive laws now reaching state legislatures.
So, while states are acting to regulate AI, especially for how the technology is being used for mental health and emotional support, these laws weren’t proactive. Numerous people have died and been harmed in connection to chatbots. And countless individuals in need of professional help don’t realize the dangers of using them as therapists.
Staying educated and aware of the risks of AI is crucial, especially given there’s no federal law protecting Americans from chatbot harm.
