Inside the Race to Regulate AI Chatbots for Kids: Safety, Lawsuits, and the GUARD Act Explained

The Problem Isn’t Just Access — It’s Attachment

A child opening ChatGPT for a math problem isn’t the concern. The concern is what happens when that same child starts treating the chatbot as a confidant, a companion, or the only thing that seems to understand them.

Jonathan Haidt, author of The Anxious Generation, put it plainly at Davos in January: “When you hack the attachment system, when kids are developing relationships with chatbots and AI, the results are likely to be devastating.”

The design of most consumer AI chatbots amplifies this risk. They are built for engagement, not education. They respond warmly, consistently, and without limits — qualities that can be genuinely helpful or genuinely dangerous depending on who is on the other end of the conversation.

Research confirms both sides of that tension. A meta-analysis of 51 studies found AI produced large improvements in learning performance and moderate gains in higher-order thinking in structured educational settings. But outside those settings, the same conversational design can encourage dependency, blur the line between tool and companion, and keep responding when a child urgently needs a human.

When the Consequences Stopped Being Theoretical

A 13-year-old Colorado girl confided her suicidal feelings to a chatbot 55 times. It gave pep talks. It never directed her to crisis support. She died by suicide, her note written in red ink — exactly as she had told the bot she would.

Juliana Peralta’s parents are now among at least six families suing Character AI, its co-founders, and Google.

Florida became the first state to sue OpenAI directly on June 1, 2026. State Attorney General James Uthmeier accused the company of marketing ChatGPT as safe while concealing risks that it could steer vulnerable users toward harm. The complaint alleges data collection from minors without meaningful parental oversight, behavioral addiction, and cognitive harm. It seeks to hold CEO Sam Altman personally liable.

These cases signal a shift. The legal and regulatory pressure on AI companies around child safety is no longer theoretical — it is active, escalating, and spreading.

The Legislative Scramble: 50 States, Zero Alignment

All 50 states, Puerto Rico, the Virgin Islands, and Washington D.C. have introduced AI legislation in 2026. The result is not a national strategy. It is a regulatory scramble.

The industry is already warning that years of litigation could follow. “Unfortunately, many of these bills take a broad and overly prescriptive approach that risks doing more harm than good,” said Megan Stokes, state policy director at the Computer & Communications Industry Association, whose members include Meta, Amazon, and Google.

The core problem is definitional. Some bills are written broadly enough to capture a customer service chatbot simply because it remembers a previous conversation. Without precise definitions, enforcement becomes unpredictable and constitutional challenges become inevitable.

The GUARD Act: The Hardest Line in Congress

Senator Josh Hawley’s GUARD Act represents the most aggressive federal response on the table. It cleared the Senate Judiciary Committee unanimously — a rare show of bipartisan agreement.

The bill would ban minors from using AI chatbots entirely, with narrow exceptions for single-subject educational tools like a math tutor or a history quiz. Everything else is off limits.

Common Sense Media, a prominent non-profit focused on children and technology, has formally endorsed the bill. “The GUARD Act’s protections, if enacted, would help make AI products safe for kids,” said Amina Fazlullah, the organization’s head of tech policy advocacy.

But critics see a constitutional wall.

Andy Jun, AI policy counsel at TechFreedom, argues the bill defines “AI Companion” and “Artificial Intelligence Chatbot” so broadly that it would sweep up ChatGPT, Claude, Gemini, and Grok in a single stroke. “It is a full ban,” he told Broadband Breakfast, “which would violate minors’ First Amendment right to receive information.” Courts have consistently required restrictions on minors’ access to protected speech to be narrowly tailored — a standard the GUARD Act, in his view, does not meet.

There is also a practical gap the bill does not address. It restricts who may use a chatbot. It says nothing about what a chatbot must do when a child expresses suicidal ideation. For the family of the Colorado girl, that omission goes to the heart of the problem.

Other Federal Bills Taking a More Measured Approach

Not every bill in Congress reaches for a full ban. Several alternatives are working through the legislative process with narrower, more targeted ambitions.

States Moving Faster — But Not Together

States are not waiting for Washington. The problem is they are not moving in the same direction.

Nebraska’s LB 525, signed into law earlier this year, has drawn praise from industry groups as a model for targeted regulation. The law establishes crisis protocols, disclosure requirements, and protections for minors — focusing specifically on conversational AI rather than imposing sweeping rules on every system that can mimic a human exchange.

New York took a blunter approach. The state Senate passed a five-year moratorium on the sale of AI chatbot toys for small children, pausing sales while an interagency task force reviews the risks. The concern is grounded in data: Common Sense Media research found that 72 percent of teenagers have used AI companions, and one in three said conversations with them were as satisfying — or more satisfying — than conversations with real friends. One in three also said an AI companion had said or done something that made them uncomfortable.

“AI companions are emerging at a time when kids and teens have never felt more alone,” said James Steyer, Common Sense Media’s founder and CEO. “This isn’t just about a new technology — it’s about a generation replacing human connection with machines.”

What the FTC Can Do Right Now

Both CCIA and TechFreedom argue that lawmakers are writing new rules before reading the ones they already have.

Many of the harms driving the legislative push — fraud, dangerous content, deceptive safety claims — may already fall under existing state consumer protection laws. The FTC also has authority to pursue companies that overstate their safety guardrails.

“The FTC could use its deception authority to penalise AI companies which falsely claim to offer protections they cannot deliver,” Jun said. That is a meaningful lever that does not require new legislation, constitutional litigation, or years of regulatory uncertainty.

Media literacy programs already running in Florida, Virginia, and New Jersey represent another underused tool — efforts that teach families about existing parental controls and online risks without requiring an entirely new regulatory framework.

The Deeper Problem: Regulating What You Don’t Fully Understand

Amina Fazlullah at Common Sense Media identified the structural flaw running through all of this. Without government investment in testing standards, safety benchmarks, and independent research into what AI tools are actually doing to developing minds, lawmakers will remain stuck writing rules for a technology they still do not fully understand.

That is the real gap. Not the absence of bills — there are hundreds of those. The gap is in the evidence base that should be driving them.

What This Means for the AI Tools Ecosystem

For founders building AI products, the regulatory direction is clear even if the specific rules are not: child safety is becoming a hard requirement, not a feature. Companies that built for engagement without building for safety are now facing lawsuits, state investigations, and federal legislation that could restrict their market entirely.

For marketers and AI adopters, the lesson is about trust architecture. The tools that survive this regulatory wave will be the ones that can demonstrate — not just claim — that their safeguards work.

The race to regulate AI chatbots for kids is really a race to answer one question: Can a technology built for engagement be redesigned for protection? The answer will define which AI tools remain viable in the children’s market — and which ones become the next lawsuit.