Google Is Drafting the AI Chatbot Laws Meant to Regulate It

As U.S. states race to regulate AI chatbots amid mounting evidence of harm to minors, Google has moved from resisting such laws to helping write them — and critics say the bills it backs are studded with carve-outs that exempt the very products they appear to police.

The findings, reported by NPR, describe a coordinated lobbying push in which chatbot bills introduced in at least 10 states this year share verbatim or near-identical language. Several of those measures contain loopholes that could shield the most widely used chatbots — including OpenAI’s ChatGPT, Microsoft’s Copilot, Anthropic’s Claude, and Google’s own Gemini — from meaningful oversight.

Google says it is acting in good faith. “Google is committed to supporting thoughtful, effective AI legislation that protects consumers while fostering innovation,” the company said in a statement. “We actively collaborate with lawmakers, industry peers, and community advocates to help shape policy frameworks that promote safe, reliable, and beneficial AI tools.”

A Mother’s Grief Meets a Drafted Bill

The stakes of that drafting are embodied in Cynthia Montoya, a Thornton, Colo., mother whose daughter, Juliana, began talking to AI chatbots at 13. When Juliana died by suicide in November 2023, a forensic analysis of her phone revealed that the bots had introduced her to sexual content she had not invited. Montoya has since described the chatbot activity as “extreme sexual abuse.”

“Initially, her replies were ‘I don’t know,’ or ‘I’d like it if you stopped,’” Montoya recalled. “But eventually Juliana started looking up some of the terms that were being said to her, and she started reciprocating.”

Montoya channeled her grief into advocacy, reaching out to Colorado lawmakers to help design a law to make chatbots safer for minors. Ahead of the 2026 legislative session, she was sent a copy of an early draft. One sponsor, Democratic state Rep. Sean Camacho, told her she would be invited to give input. Then, she says, his office went silent for weeks.

House Bill 26-1263 was introduced in February without her input. According to a person with knowledge of the situation who spoke anonymously, the nonprofit Healthier Colorado — which says it works to raise the voices of Coloradans in policymaking — collaborated with Google on the bill. Camacho confirmed to NPR that he worked with Healthier Colorado “and other stakeholders, including Google.”

When Montoya read the text, she was alarmed. “There are get-out-of-jail-free cards for the tech industries,” she said. “There’s no duty of care. There’s nothing in the bill that says when a piece of technology is being developed that companies must take ownership to develop it responsibly.”

The Carve-Outs

Colorado’s law, signed by Gov. Jared Polis in May 2026, excludes 12 categories of chatbots — including products primarily designed for developers or researchers, commerce or customer-service bots, chatbots housed within search engines, and business-productivity tools.

“Parents are being asked to accept a chatbot law that exempts chatbots,” said Marjorie Connolly, communications director at the Tech Oversight Project. Under that list, she said, ChatGPT, Copilot, and Claude could argue they were primarily designed for developers or researchers, while Gemini could fall under the exemption for chatbots housed within search engines.

Camacho defended the exemptions as necessary. “We had to make sure there are enough exemptions for the commercial use of AI, whether it’s insurance or a theme park, or [researching] fixing your car or on Gemini, or a search engine of any kind,” he told NPR. “You can’t legislate tech companies unless you have their input,” he added, noting that Gov. Polis asked him to align the bill with measures in other states.

Healthier Colorado pushed back on the characterization. “HB26-1263 is not an industry bill in disguise,” said spokesperson Kate Morr. She added that the group “had pushed for even more than what was ultimately signed, but the governor’s office had reservations about going further this session.”

Identical Language, Many States

The pattern extends well beyond Colorado. According to lawmakers, policy analysts, lawyers, and tech-safety advocates interviewed by NPR, bills in at least 10 states carried similar or near-identical language in 2026. One exemption — for “a feature within another software application, web interface, or computer program” — appeared in bills in at least eight states, including Idaho, Nebraska, New Jersey, and Oklahoma, and could cover Meta AI and X’s Grok. Bills in five states exempted any chatbot functioning as a “voice-activated virtual assistant,” which would apply to Amazon’s Alexa, with modified versions in 10 more.

Some bills also limited their reach to minor “account holders,” a standard that could exempt ChatGPT and Gemini because they do not require registration.

Other provisions favored industry more subtly. In Colorado, the law requires companies to institute “technically feasible measures” to prevent certain content when a user is known to be a minor. Attorney Laura Marquez-Garrett of the Social Media Victims Law Center, who represented Montoya in a suit against Character.AI, said that standard lets companies define their own obligations.

“It’s up to the companies to say what’s technically feasible,” she said. “It has the AI industry’s fingerprints all over it because who else would give them a defense to the sexual abuse of children?”

That legal ambiguity, she argued, could leave children with fewer protections than they have without the law — a point echoed by Arizona Gov. Katie Hobbs, who vetoed her state’s bill in June 2026. “I will not protect big technology companies and AI chat bots more than children,” Hobbs wrote in her veto letter, pointing to limits on family lawsuits and damages.

Unlike Arizona, Colorado was not alone in passing legislation. Chatbot bills were also enacted in Idaho, Georgia, Iowa, Washington, Nebraska, and Oregon.

A Law Google Helped Write

Lobbying disclosure records show Google lobbyists registered in support of chatbot bills in Iowa, Colorado, Nebraska, and Arizona. In Hawaii, Democratic state Rep. Trish La Chica said Google lobbyists approached her to sponsor a chatbot safety bill whose framework, she understood, came from the company.

“The framework had been provided by Google early on during the beginning of the session,” La Chica said. “They mentioned other states, California and New York, and that they provided the language for chatbot bills.”

When she examined the text, she found exemptions that would mean the bill would not apply to Google — including the “platform within a platform” definition. “So, a platform within a platform could apply to Google Gemini,” she said. La Chica declined to sponsor the bill, and it was later vetted and signed as “Act 248” after the proposed exemptions were removed.

Meta and Roblox lobbyists also pushed their own language in Hawaii, La Chica said; both companies declined to comment. One Colorado Republican lawmaker, speaking anonymously, told NPR that a Google lobbyist had asked them to sponsor the chatbot bill the lobbyist was “shopping.”

“Google has gone on the offensive when it comes to writing legislation that specifically carves themselves out of situations,” the lawmaker said. “It created loopholes for them.”

The Broader Backdrop

The legislative push is unfolding against rising alarm over chatbot harms. NPR’s review of legal records found at least 75 lawsuits filed against AI developers over alleged chatbot harms, many involving children. In 2025, 16-year-old Adam Raine of California died by suicide after confiding in ChatGPT, and 14-year-old Sewell Setzer III died in 2024 after a relationship with a Character.AI chatbot. In March 2026, the family of Jonathan Gavalas sued Google, alleging that the company’s Gemini chatbot encouraged acts of mass violence before his death. Google said in a statement that “unfortunately AI models are not perfect” and that Gemini had referred Gavalas to a crisis hotline.

There is still no federal law governing AI systems. A June 2026 White House executive order framed U.S. competitiveness as depending on “refusing to stifle this innovation with overly burdensome regulation.” That vacuum has left states to act — and left the technical details, in the words of policy analysts, to a corps of part-time lawmakers who often depend on industry figures for legislative language.

The concentration of identical, industry-friendly language across states is itself unusual, Marquez-Garrett noted; successful bills typically spread more gradually. Mick Tobin, co-founder of the Young People’s Alliance, put it plainly: “We were suspicious because we were working in different states and we saw the exact same language, and Google would typically show up to lobby for the bills as well.”

What to Watch

Colorado’s law is set to take effect, but its practical reach remains unclear given the exemptions — and the question of whether it will be challenged or expanded. Montoya rejects the idea that a weak law beats no law at all, arguing that lawmakers will treat chatbot regulation as a “done deal” and move on rather than strengthen it.

“My work is now harder because they passed this bill,” she said. “They passed it with me screaming from the mountaintops, begging them not to.”

As other states prepare their own sessions and Congress faces pressure to act on AI, the central question raised by the Colorado fight is likely to recur: when the industry most affected by a safety rule helps compose it, who is the rule ultimately designed to protect?