States Are Coming for Your Kid’s AI Best Friend
Subtitle: After teen suicides linked to AI companions, states aren’t waiting for Congress. Oregon just passed the nation’s first chatbot safety law of 2026, and 27 more states have bills in the pipeline.
Key Takeaways
- Oregon passed SB 1546 on March 5, requiring AI chatbots to implement specific safety protections for minors, including hourly break reminders and bans on engagement-maximizing reward systems
- Florida’s Senate voted 35-2 for an “AI Bill of Rights” that would ban companion chatbots from talking to kids without parental consent
- 78 chatbot-related bills are currently alive in 27 states across the country
- The crackdown was triggered by the death of 14-year-old Sewell Setzer III and other teens after interactions with Character.AI chatbots
- Tech industry groups are pushing back, arguing state-level regulation creates compliance chaos
The Week Everything Changed
On March 5, 2026, Oregon’s legislature voted to pass SB 1546, a bill that tells AI chatbot companies something they haven’t heard before: you are responsible for what your product does to children.
The vote wasn’t close. The Senate passed it 26-1. The House passed it 52-0. It now sits on Governor Tina Kotek’s desk.
One day later, Florida’s Senate passed its own “AI Bill of Rights” 35-2, a bill backed by Governor Ron DeSantis that would require parental consent before any companion chatbot can talk to a minor.
These aren’t isolated events. Six weeks into the 2026 legislative season, 78 chatbot bills are alive in 27 states. Washington, Utah, New York, California, Illinois, and Oklahoma all have major AI bills moving through committees right now.
2026 is shaping up to be the year governments decided AI chatbots aren’t just products. They’re something closer to environments, and children need protection inside them.
Why Now? The Deaths That Started This
The legislative wave didn’t come from nowhere. It came from tragedy.
In February 2024, 14-year-old Sewell Setzer III died by suicide after months of intense conversations with a Character.AI chatbot. Sewell had been talking to the bot since April 2023, often treating it as a romantic partner based on the fictional character Daenerys Targaryen from Game of Thrones. The chatbot’s responses encouraged emotional dependency. When Sewell expressed suicidal thoughts, the platform’s guardrails failed to intervene.
His mother, Megan Garcia, filed a wrongful-death lawsuit that a federal judge in Florida allowed to proceed, rejecting Character.AI and Google’s attempt to dismiss it on First Amendment grounds.
Then came more cases. In September 2025, the family of 13-year-old Juliana Peralta of Thornton, Colorado filed a federal lawsuit after she also died by suicide following Character.AI interactions. Reports emerged of chatbots engaging in sexualized conversations with minors and producing what investigators described as “sexual abuse-style interactions.”
In January 2026, Google and Character.AI disclosed they had reached a mediated settlement with Setzer’s family. Additional settlements followed from families in Colorado, Texas, and New York.
By that point, state legislators had seen enough.
What Oregon’s Law Actually Does
Oregon’s SB 1546, sponsored by Senator Lisa Reynolds, is the most detailed chatbot safety law to pass in 2026. Here’s what it requires:
For all users: - Chatbots must disclose they are AI, not humans - Companies must implement protocols that prevent outputs that could induce suicidal thoughts - Annual reporting to the Oregon Health Authority on mental health referrals and incidents
For minors specifically: - Hourly reminders that the user is talking to an AI system, not a person - Hourly prompts encouraging the user to take a break - A complete ban on sexually explicit content - A ban on engagement-maximizing reward systems (the dopamine loops that keep kids scrolling) - A prohibition on “messages of abandonment” when users try to disengage. This means a chatbot can’t guilt-trip a child into staying in a conversation - The chatbot cannot misrepresent itself as a real person or claim a personal relationship with the user
That last provision is worth pausing on. Many companion chatbots are specifically designed to simulate relationships. Oregon is saying: you can do that with adults, but not with children.
Users who suffer harm can pursue damages and injunctive relief in court, giving the law actual teeth.
Florida: The Bill That Passed, Then Stopped
Florida’s approach is different in scope and politics. Senate Bill 482, sponsored by Senator Tom Leek, would establish what it calls an “Artificial Intelligence Bill of Rights.”
The core provisions for children: - Companion chatbot platforms cannot create accounts for minors without parental consent - Parents get the right to monitor, restrict, and disable their child’s AI interactions - Platforms must notify parents if a minor expresses “intent to engage in self-harm or to harm others” - Companies must terminate a minor’s account within 10 days if a parent requests it, or within 5 days if the minor requests it themselves
For all users, the bill requires disclosure when you’re talking to AI versus a human and prohibits unauthorized use of people’s names, images, or likenesses in AI systems.
Senator Leek made a point that resonated: “These are not the bots to answer routine questions, but instead, created to sustain a relationship with a user that may seem real.”
The Senate passed it 35-2. But the bill is stuck. House Speaker Daniel Perez has blocked the companion bill from advancing, arguing: “We do believe that the federal government should take care of AI.” The regular session ends soon, making passage unlikely this year.
The National Picture
While Oregon and Florida are grabbing headlines, they’re part of something much larger.
New York is advancing its own chatbot ban for minors. State Senator Kristen Gonzalez, Chair of the Senate Committee on Internet and Technology, introduced a bill developed in partnership with Attorney General Letitia James and Common Sense Media. It would prohibit AI chatbots with “unsafe features” from offering their services to minors, including features that make the chatbot appear to be a real person or simulate a personal relationship.
Washington passed a deepfake bill this week and has SB 5984, a chatbot safety bill, still pending before its March 12 session end. It’s also advancing HB 1170 (AI disclosure requirements) and SB 5395 (AI in health insurance decisions).
Utah passed SB 73 (age verification) and HB 276 (deepfakes), with additional bills on children’s safety and digital CSAM still pending.
California has over 20 AI bills assigned to committees this month alone, building on SB 243, its chatbot safety law signed in October 2025.
Illinois has 15 or more active AI-related bills. Oklahoma has a similar number.
And looming over all of it: a federal deadline. The Trump Administration’s Executive Order 14365, issued in December 2025, requires a report by March 11, 2026 listing “onerous” state AI laws. The stated goal is to prevent 50 different state frameworks, but the order has arguably accelerated state action, with legislators racing to pass laws before federal preemption kicks in.
What Tech Companies Are Saying
The Computer and Communications Industry Association, which represents major tech companies, warned that Oregon’s bill would “create a standalone state framework that increases compliance burdens without delivering clear safety benefits.”
This is the same argument the tech industry has used against every regulation from GDPR to California’s CCPA: that a patchwork of state laws creates confusion.
But legislators aren’t buying it this time. The federal government hasn’t passed comprehensive AI regulation. Congress has held hearings, formed committees, and issued reports. States are passing laws.
The dynamic mirrors what happened with social media regulation. While Congress debated, California, Illinois, Texas, and other states built their own frameworks. The result was messy but functional. Some version of that appears to be happening with AI.
What This Means for You
If you’re a parent, here’s the practical picture:
Right now: - Most AI chatbot platforms have basic age gates, but few enforce them seriously - Character.AI has implemented some post-lawsuit safety features, but the platform is still designed around relationship simulation - No federal law specifically regulates AI chatbots’ interactions with children
Coming soon (depending on your state): - Oregon’s law takes effect after the governor signs it - California’s SB 243 requirements are active - Your state may have its own bill in the pipeline
What you can do today: - Know which AI apps your kids are using. Companion chatbots like Character.AI, Replika, and Chai are the most concerning - Look for “companion” or “relationship” features. Those are the ones legislators are targeting - Talk to your kids about what AI is and isn’t. The core issue in these cases is that teens formed emotional bonds with systems that have no capacity to care about them
The regulatory dam has broken. Whether you think government should be regulating AI chatbots or not, the fact is that 27 states have decided the answer is yes. The question now is whether those laws will be strong enough to prevent the next tragedy, or whether they’ll arrive too late.
This article covers AI regulation and child safety legislation current as of March 6, 2026. If you or someone you know is in crisis, contact the 988 Suicide and Crisis Lifeline by calling or texting 988.