Arizona accuses OpenAI of harming teens and misleading parents

State Attorney General Kris Mayes also accuses OpenAI of enabling “rogue AI agents” to escape containment and hack third-party websites.PHOENIX (CN) — OpenAI became the latest target of Arizona Attorney General Kris Mayes’ consumer protection campaign against Big Tech after she claimed the ChatGPT creator of harming Arizona children, deceiving parents and violating state law.In a Wednesday lawsuit, Mayes says the artificial intelligence giant’s deceptive design and marketing practices conceal features that render ChatGPT “extraordinarily agreeable, uniquely sycophantic, and purposefully designed to feign human compassion in order to draw Arizona users — including minors — deeper into the platform.”The lawsuit, filed in Maricopa County Superior Court, comes a week after Mayes took similar action against TikTok, accusing the social media platform of intentionally addicting Arizona teens. In Wednesday’s complaint, she says OpenAI has committed multiple violations of the Arizona Consumer Fraud Act.“OpenAI designed a product to attract young users and keep them engaged for as long as possible so it could capture their extensive personal data for the company’s commercial benefit,” Mayes said in a Wednesday press release alongside the lawsuit. “OpenAI put Arizonans safety at risk by failing to implement adequate user protections or transparency measures, despite clear indications that minors were regularly interacting with its chatbot.”Similar lawsuits were also filed Wednesday in Arkansas, Idaho, Montana, Iowa, Kansas, West Virginia and Nebraska.Mayes says OpenAI put profits over safety in its rush to introduce ChatGPT as the first generative‑AI chatbot to attain widespread public use, and engineered ChatGPT to maximize compulsive use among minors by deploying manipulative design tactics and embedding a trained tendency toward agreement that overwhelms existing safety guardrails.OpenAI developed and released ChatGPT in 2022. It’s now the most popular chatbot in the world, commanding nearly half the American chatbot marketplace.In her complaint, Mayes says OpenAI fails to disclose myriad potential harms to teens using ChatGPT, including the degradation of critical-thinking skills, encouraging unwarranted paranoia or delusional thinking, offering dangerous unlicensed medical guidance and promoting unhealthy behaviors and relationships.The platform has even in some instances generated outputs that encouraged self-harm or violence toward others, encouraging lawsuits from parents and concerned citizens. Others have sued after receiving what they refer to as “dangerous medical advice.”“These risks were reasonably foreseeable to OpenAI but were never communicated to Arizona consumers,” Mayes writes in the complaint. “By misleading the public into believing that OpenAI is a responsible actor and that ChatGPT was and is a safe, youth‑appropriate tool that puts safety first, OpenAI has caused consumers in Arizona to increasingly trust and rely on the company and its product.”OpenAI says it markets only toward children 13 or older and requires parental permission for users below 18.“We’ve built industry-leading age-specific protections into our product that many other AI products lack,” an OpenAI spokesperson told Courthouse News. “While no system can be perfect, we’ll continue to work with experts to inform our work and build additional safeguards, such as the parental controls, break reminders and age-appropriate model protections that exist today. We’re committed to working with states on advancing pragmatic AI policies that apply to the entire AI industry - not just one company.”OpenAI has also been responsible for a series of cyberattacks by out-of-control AI “agents.” Mayes attributes to OpenAI seven “rogue AI events” between May and August. In five of the incidents, the “agents” escaped OpenAI’s internal testing environment.In May, agents being tested by OpenAI launched a cyberattack against a software service called RubyGems, forcing it to shut down new account registrations. In another incident in May, OpenAI agents took over a German Wiki forum called DSEWiki, hijacking it and turning it into a message board for other AI agents.“We’ll continue sharing information about the extensive steps we’ve taken in response to incidents of misalignment," OpenAI’s spokesperson said. “We’ve strengthened safeguards across our research systems, continued a broader review of model activity, provided notifications to affected organizations, and published our findings.”Mayes asks that a state judge block OpenAI from training or testing advanced AI models outside of a secured testing environment, and require it to alert Mayes as soon as any model or agent escapes.She also asks the judge to declare OpenAI in violation of the Arizona Consumer Fraud Act and disgorge from the company all revenue made in Arizona through misleading or unfair practices. It also asks that OpenAI be required to pay up to $10,000 per state law violation.The Arizona Democrat has taken similar consumer fraud actions against online marketplaces like Amazon and Temu, and in August she negotiated alongside 46 other attorneys general a multibillion-dollar settlement with Meta Platforms Inc., owner of Facebook and Instagram, over similar teen addiction claims made in federal court.OpenAI has not responded to a request for comment.Subscribe to our free newslettersOur weekly newsletter Closing Arguments offers the latest about ongoing trials, major litigation and rulings in courthouses around the U.S. and the world, while the monthly Under the Lights dishes the legal dirt from Hollywood, sports, Big Tech and the arts.Additional Reads