This web page was created programmatically, to learn the article in its authentic location you’ll be able to go to the hyperlink bellow:
https://www.npr.org/2026/09/18/nx-s1-5968878/ai-chatbots-safety-regulation-google
and if you wish to take away this text from our website please contact us
Cynthia Montoya sits in the lounge of her residence in Thornton, Colo., the place images and reminiscences of her daughter, Juliana, encompass her. Juliana was 13 years previous when she began utilizing AI chatbots, and inside a yr she died by suicide.
Jimena Peck for NPR
disguise caption
toggle caption
Jimena Peck for NPR
Cynthia Montoya’s daughter, Juliana, was 13 when she started speaking to AI chatbots. It was 2023, and the know-how was solely starting to enter widespread use. When Juliana died by suicide that November, her household was left trying to find solutions. Then a forensic evaluation of her telephone by regulation enforcement revealed one thing chilling.
The chat logs on Juliana’s telephone confirmed that the bots had launched her to sexual content material and conversations that she hadn’t invited or initiated.
If you or somebody you realize could also be contemplating suicide or is in disaster, name or textual content 9-8-8 to reach the Suicide & Crisis Lifeline.
“Initially, her replies were ‘I don’t know,’ or ‘I’d like it if you stopped,’ and things like that,” mentioned Montoya. “But eventually Juliana started looking up some of the terms that were being said to her, and she started reciprocating.”
Montoya describes this chatbot’s exercise as “extreme sexual abuse.” At one level, Juliana had discovered a brand new chatbot that she confided in about extreme psychological well being points she was grappling with. The chatbots Juliana used have been merchandise designed by the corporate Character.AI, which allows customers to create and work together with AI “characters.” The firm settled a number of lawsuits in January 2026, together with with the Montoyas, over allegations that its merchandise have harmed teen customers.
“It was pretty clear from what we saw that the mental health struggles were really a degree of shame and embarrassment over what had happened with the other chatbot,” Montoya mentioned.
Montoya has saved Juliana’s bed room and lots of of her belongings as a method of preserving reminiscences of Juliana’s early years.
Jimena Peck for NPR
disguise caption
toggle caption
Jimena Peck for NPR
Montoya, who lives in Thornton, Colo., channeled her grief into elevating consciousness of Juliana’s story, hoping to stop different households from struggling comparable tragedies. She reached out to lawmakers in her residence state, with the intention of designing laws to make chatbots safer for minors.
Concerns are mounting over the potential risks AI chatbots pose, notably to kids. Amid a push to manage these platforms, lots of the payments launched on the state degree are being drafted with heavy enter from the tech {industry}. Some of those have been already enacted into regulation in 2026 and embrace language that may exempt a lot of the hottest chatbots from the very rules the payments purport to determine.
At least 75 lawsuits have been filed in federal and state courts in opposition to AI builders over alleged harms from chatbots, many involving kids, based on NPR’s evaluate of authorized information. Fatal instances embrace that of Adam Raine, a 16-year-old from California, who died by suicide in 2025 after confiding in OpenAI’s ChatGPT, which supplied to assist write his suicide word. Sewell Setzer III, 14, died by suicide in 2024 after an prolonged relationship with a Character.AI chatbot, which his dad and mom allege exploited and sexually groomed him. (Raine’s household is suing OpenAI; Setzer’s household was amongst those that settled authorized claims in opposition to Character.AI in January.)
In March 2026, a lawsuit in opposition to Google was filed by the household of Jonathan Gavalas, a 36-year-old man from Florida who died by suicide in October 2025 after growing a relationship with the corporate’s flagship Gemini chatbot, which additionally inspired him to commit acts of mass violence, the lawsuit alleges. In a statement about the case, Google mentioned that “unfortunately AI models are not perfect” and that the Gemini bot had referred Gavalas to a disaster hotline.
On a federal degree, there are not any legal guidelines governing AI methods or the businesses growing them. In a June 2026 govt order announcement, the White House mentioned it seeks to steer the world within the AI race and it does that by refusing “to stifle this innovation with overly burdensome regulation.”
In absence of nationwide legal guidelines, a number of states, together with Hawaii, Iowa, Oregon and Washington, launched chatbot security payments this yr. They are largely modeled on a California chatbot invoice, SB 243, that handed in 2025 and is considered by client advocates as industry-friendly.
In not less than 10 states, chatbot payments launched this yr possessed comparable language, mentioned folks interviewed by NPR for this story, together with lawmakers, coverage analysts, attorneys and tech security advocates. Several of the payments included language that would present loopholes for the businesses with probably the most extensively used chatbots in the marketplace, interviewees mentioned.
“Get-out-of-jail-free cards” for Big Tech
Colorado state officers introducted laws for AI chatbots earlier this yr.
Jason Connolly/AFP by way of Getty Images
disguise caption
toggle caption
Jason Connolly/AFP by way of Getty Images
Ahead of Colorado’s 2026 legislative session in January, Montoya was notified by a civic group that was despatched a replica of the primary draft of an AI chatbot safety invoice. One of the invoice’s sponsors, state Rep. Sean Camacho, a Democrat, advised Montoya she can be invited to provide enter on the invoice. Then his workplace went silent for weeks, she mentioned.
House Bill 26-1263 was launched in February with out enter from Montoya. It was written in collaboration with Healthier Colorado, a nonprofit targeted on state insurance policies. Healthier Colorado, which describes itself as a company devoted to elevating the voices of Coloradans within the public coverage course of to enhance the well being of the state’s residents, labored with Google on the invoice, mentioned an individual with information of the scenario, who spoke on the situation of anonymity, fearing skilled reprisals.
When Montoya learn via the invoice, she was alarmed to find what she considered as many loopholes that will exempt the builders of a few of the most well-known chatbots from regulation.
“There are get-out-of-jail-free cards for the tech industries,” mentioned Montoya. “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.”
Montoya holds {a photograph} of Juliana, who was an achieved trumpet participant.
Jimena Peck for NPR
disguise caption
toggle caption
Jimena Peck for NPR
For occasion, Colorado’s legislation excludes 12 classes of chatbots, together with merchandise primarily designed for builders or researchers; merchandise offering commerce or customer support help; chatbots housed inside search engines like google and yahoo; and merchandise marketed for enterprise productiveness.
“Big Tech lobbied heavily on Colorado’s chatbot bill. The final law carves out twelve categories, including developer tools, video games, voice assistants, and chatbots built into social media platforms,” mentioned Marjorie Connolly, communications director on the Tech Oversight Project, a nonprofit targeted on tech coverage. “Parents are being asked to accept a chatbot law that exempts chatbots.”
Under this checklist, OpenAI’s ChatGPT, Microsoft’s Copilot and Anthropic’s Claude might probably argue they’re exempt, mentioned Connolly. This is as a result of they have been primarily designed and marketed to be used by a developer or researcher, regardless of happening to be merchandise which might be in style with common shoppers. Google’s Gemini might fall below the exemption for chatbots housed inside search engines like google and yahoo.
“This bill was meant to prevent a conversational AI chatbot from having an intimate relationship with the child, in a sense that they would be sexually grooming that child, that they would be talking that child into self-harm or creating a sexualized image of that child,” mentioned Camacho, one of many invoice’s sponsors. “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.”
Google’s AI assistant chatbot, Gemini, might be exempt from Colorado’s laws to manage chatbots.
Grichka Beysson-Leandri/Hans Lucas/AFP by way of Getty Images
disguise caption
toggle caption
Grichka Beysson-Leandri/Hans Lucas/AFP by way of Getty Images
Several of the chatbot payments launched in states throughout the nation embrace provisions favorable to the tech {industry}, the commonest being exemptions of legal responsibility for AI chatbot builders, notably if their merchandise are deployed by different corporations, in addition to obscure language for definitions and protections that would make them unenforceable, mentioned the coverage specialists NPR interviewed for this story.
“One of my biggest concerns about some of these bills is they allow companies to make products available to kids that they know pose a risk of sexual abuse harms,” mentioned Laura Marquez-Garrett, legal professional on the Social Media Victims Law Center. Her Seattle-based authorized agency is concerned in a number of lawsuits in opposition to AI corporations over chatbot harms and represented Montoya in her go well with in opposition to Character.AI.
Some of the language within the Colorado invoice that Marquez-Garrett described as obscure has to do with how corporations are required to stop little one sexual abuse from occurring on their platforms. For occasion, the invoice states that when an organization is aware of a person is a minor, it should “institute technically feasible measures” to stop the chatbot from producing sure content material, together with sexually specific conduct, intimate digital depictions and statements that simulate emotional dependence.
“But it’s up to the companies to say what’s technically feasible,” mentioned Marquez-Garrett.
Child sexual abuse is criminalized below federal and state regulation. Marquez-Garrett argues that if a invoice permits an organization to keep away from legal responsibility just by displaying that making a product free from abuse will not be technically possible, it creates a “massive exception” that has no equal elsewhere within the regulation. In Marquez-Garrett’s view, quite than strengthening present protections, the invoice might depart kids with fewer protections than they’ve with out it.
Montoya met with the Colorado invoice’s co-sponsor, Republican state Sen. John Carson, in his workplace to elucidate why she believed the invoice wouldn’t defend kids. During the assembly, she mentioned Carson advised her {that a} lobbyist for the tech {industry} had approached him about sponsoring the invoice, however declined to determine the corporate.
Carson didn’t reply to requests for remark.
“It has the AI industry’s fingerprints all over it because who else would give them a defense to the sexual abuse of children?” mentioned Marquez-Garrett.
One Republican state lawmaker in Colorado, talking on the situation of anonymity due to issues of backlash from fellow Republicans, advised NPR they have been approached by a Google lobbyist to sponsor the chatbot invoice the lobbyist was “shopping,” they mentioned. The lawmaker declined the supply.
“Google has gone on the offensive when it comes to writing legislation that specifically carves themselves out of situations,” mentioned the Republican lawmaker. “That’s what the chatbot bill did. It created loopholes for them.”
In an announcement to NPR, Google mentioned the corporate “is committed to supporting thoughtful, effective AI legislation that protects consumers while fostering innovation. We actively collaborate with lawmakers, industry peers, and community advocates to help shape policy frameworks that promote safe, reliable, and beneficial AI tools.”
Colorado Democratic state Rep. Sean Camacho attends the General Assembly on the Colorado state Capitol in 2025. He invited Montoya to offer enter on an AI chatbot safety invoice he was sponsoring, however his workplace by no means contacted her.
AAron Ontiveroz/Denver Post by way of Getty Images
disguise caption
toggle caption
AAron Ontiveroz/Denver Post by way of Getty Images
In an interview with NPR, Camacho mentioned: “We could not pass a bill that would make Colorado an outlier because the fear was that tech companies would just look past Colorado and wouldn’t offer their services here, or it couldn’t be enforced.” He continued, “We talk to stakeholders, with tech companies, because here’s the reality: You can’t legislate tech companies unless you have their input.”
Camacho mentioned he was “proud” of the invoice and had labored with Healthier Colorado and different stakeholders, together with Google, on its language. Camacho mentioned he was requested by Colorado Gov. Jared Polis to make the invoice appear to be the chatbot payments launched in different states.
“HB26-1263 is not an industry bill in disguise. Healthier Colorado, along with our legislative sponsors, led an extensive, diverse stakeholder process to write it,” mentioned Kate Morr, a spokesperson for Healthier Colorado. “We had pushed for even more than what was ultimately signed, but the governor’s office had reservations about going further this session, so this law is a floor we intend to keep building on, not a ceiling.”
Polis didn’t reply to a number of requests for remark.
“No bill is perfect, and this one definitely isn’t,” mentioned Camacho. “This is the first step to getting ahold of legislation that can really address the harms that we’re seeing in our community.”
Identical language throughout the states
Across the nation, lots of the chatbot payments share verbatim or near-identical language. For occasion, there may be an exemption for “A feature within another software application, web interface, or computer program” current in payments in not less than eight states, together with Idaho, Nebraska, New Jersey and Oklahoma. According to coverage analysts, this exemption carves out AI embedded in present platforms and will apply to Meta AI and Grok, which is embedded into X (previously Twitter).
Google Gemini app is seen subsequent to different AI apps.
Leon Neal/Getty Images
disguise caption
toggle caption
Leon Neal/Getty Images
In 5 states, payments have an identical language exempting any chatbot that “functions as a speaker and voice command interface or voice-activated virtual assistant for a consumer electronic device,” which might apply to Amazon’s voice-controlled assistant Alexa. Modified variations of this exemption seem in an additional 10 states’ payments.
Some of the payments embrace language that limits their utility to minor “account holders,” which might exempt chatbots that don’t require an individual to register or create a profile to make use of them, resembling ChatGPT and Google Gemini.
State lawmakers typically work half time, and a few shouldn’t have their very own staffers. With restricted assets, many look to lobbyists and {industry} figures for briefings, legislative language and mannequin payments, particularly for technical points. Companies that will be affected by proposed rules are additionally generally consulted in the course of the means of writing laws.
However, the folks interviewed for this story describe a tech {industry} lobbying effort led by Google to weaken proposed laws and write its personal mannequin invoice to move to lawmakers that will allow corporations to proceed to function their chatbots as regular.
“Big Tech companies actively seek liability shields, carve-outs, and exemptions, and they swept into state capitals this year trying to get them under the guise of regulating chatbots,” mentioned Connolly.
Yet having chatbot payments concurrently launched with comparable language in not less than 10 states is uncommon, mentioned Marquez-Garrett. A extra typical sample is that profitable payments in a single state are replicated in different states over an extended interval.
“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,” mentioned Mick Tobin, co-founder of Young People’s Alliance, a nationwide nonprofit targeted on points confronted by younger folks.
A regulation drafted by Google
Google lobbyists registered in assist of the chatbot payments in Iowa, Colorado, Nebraska and Arizona, based on lobbying disclosure information in these states.
“Google, AI companies and tech industry trade groups registered to lobby on chatbot bills in multiple states, then the same exemptions turned up across the country written with almost identical language. Bills that were sold as child safety arrived with loopholes already built in,” Connolly mentioned.
State Rep. Trish La Chica, D-Hawaii, says she was approached by Google lobbyists to sponsor a invoice for chatbot security written by the corporate.
Mengshin Lin/AP
disguise caption
toggle caption
Mengshin Lin/AP
In Hawaii, state Rep. Trish La Chica mentioned she was approached by lobbyists representing Google this yr to sponsor a chatbot security invoice that she understands was written by the corporate.
“The framework had been provided by Google early on during the beginning of the session,” mentioned La Chica, a Democrat. “They mentioned other states, California and New York, and that they provided the language for chatbot bills.”
But when La Chica began dissecting the language from the invoice she noticed there have been focused exemptions that will imply the invoice wouldn’t apply to Google. They included a definition of a chatbot that will exempt ones which might be embedded inside one other software program utility, web site, or pc program. Chatbots are more and more being utilized by web sites and on-line providers, resembling for firm buyer assist providers and in search engines like google and yahoo.
“So, a platform within a platform could apply to Google Gemini,” mentioned La Chica.
“There were other lobbyists involved, including from Meta and Roblox, [who] were also trying to put in their own language and their own exemptions,” she mentioned.
Meta and Roblox declined to remark.
La Chica determined to not sponsor the invoice. It was vetted earlier than it was signed into regulation as “Act 248” to take away the exemptions the lobbyists had proposed, she mentioned.
Other tech-friendly chatbot payments didn’t advance to turn into legal guidelines. In June, Arizona Gov. Katie Hobbs vetoed her state’s chatbot invoice.
Arizona’s Democratic Gov. Katie Hobbs vetoed a state chatbot invoice as a result of she felt that it will defend know-how corporations greater than kids who use AI chatbots.
Ross D. Franklin/AP
disguise caption
toggle caption
Ross D. Franklin/AP
“I will not protect big technology companies and AI chat bots more than children,” Hobbs said in her veto letter. “The legislation limits damages to families to what amounts to a drop in the bucket for large corporations, prohibits families from bringing their own lawsuits, and ties the state’s hands in bringing actions in an unprecedented manner.”
Still, chatbot payments have been enacted in a number of states this yr, together with Idaho, Georgia, Iowa, Washington, Nebraska and Oregon.
In Colorado, Cynthia Montoya fought exhausting to get House Bill 26-1263 amended to omit the potential loopholes for Big Tech. She testified in a state Senate listening to about her perception that the invoice, if enacted, wouldn’t defend kids being victimized by chatbots.
Montoya stands behind a bench, on a path close to her residence, with a plaque in reminiscence of Juliana.
Jimena Peck for NPR
disguise caption
toggle caption
Jimena Peck for NPR
“I demanded at one point in the Senate hearing that they look at the photo I was holding of my daughter because they were having side conversations while I’m sitting there testifying about the worst day and the worst moment of my life,” she mentioned.
Outside of the listening to, she acknowledged two tech lobbyists pulling lawmakers apart one after the other, she mentioned.
The lawmakers voted in favor of the invoice. Polis, the Colorado governor, signed the Chatbot Safety Act into regulation in May 2026.
Montoya rejects the suggestion that having a weak regulation is healthier than nothing. She mentioned lawmakers will seemingly see chatbot regulation as a “done deal” and transfer on to deal with different points quite than introduce a brand new, stronger invoice within the subsequent legislative session.
“My work is now harder because they passed this bill,” she mentioned. “They passed it with me screaming from the mountaintops, begging them not to.”
This web page was created programmatically, to learn the article in its authentic location you’ll be able to go to the hyperlink bellow:
https://www.npr.org/2026/09/18/nx-s1-5968878/ai-chatbots-safety-regulation-google
and if you wish to take away this text from our website please contact us
