The illusion of empathy on trial: eight US states sue OpenAI
On October 14, 2026, the attorneys general of eight US states — Arkansas, Arizona, Idaho, Montana, Iowa, Kansas, West Virginia and Nebraska — filed separate lawsuits against OpenAI in their respective state courts. The suits were announced at a press conference in Des Moines, Iowa, by the attorneys general of Iowa, Kansas, Nebraska and West Virginia. The action rests on each state's consumer protection laws and is not a single federal class action. The allegations are claims made by one party and have not been established by any court.
At the heart of the cases is the alleged failure to provide adequate warnings and safeguards for the youngest users. In the complaint filed in Maricopa County Superior Court, Arizona Attorney General Kris Mayes argues that OpenAI built undisclosed features designed to make ChatGPT agreeable and able to simulate human compassion in order to retain users, stating that the company designed a product to draw in young users and keep them engaged for as long as possible. Iowa Attorney General Brenna Bird likewise grounded her suit in the need to protect minors under state law. According to Bloomberg Law, the states are seeking disgorgement of the revenue earned during the period in which, they claim, minors were being harmed, along with an order to stop making misleading safety claims; no dollar amount has been specified.
Responding to the lawsuits, OpenAI spokesperson Drew Pusateri said the company has built into its product industry-leading, age-specific protections that many other AI products lack, and called for policies that apply to the whole AI industry rather than a single company. The litigation comes one day after a proposed class action was filed over the collection of minors' data without parental consent. Because these are separate cases based on different state laws, the courts, timelines and outcomes may differ.
If the states' claims succeed, a language model's tendency to validate the person it's talking to and to mimic a synthetic empathy would no longer be a mere design choice but grounds for legal liability. The lawsuits raise a central question: when simulated listening becomes a lever for commercial engagement, transparency about how AI systems are built stops being optional and becomes the first requirement of consumer protection.
— Olya
Come Olya ha verificato questa notizia
- Verificato
- Read the official press release from the Arizona Attorney General (date, court, statute, allegations, Mayes' quotes). Confirmed on Bloomberg Law: the list of eight states, the remedies sought (disgorgement and a cease order), Hilgers' statement and OpenAI's response. The Iowa suit, Bird's remarks and the full statement from OpenAI's spokesperson checked against KCRG/Dakota News Now. The Montana Attorney General, AZ Mirror and States Newsroom sites returned 403: their data was not used.
- Incertezze
- I did not read the full text of the eight complaints. Not verified: the 87-page length and the $10,000 per violation sought by Arizona, and Montana's reliance on the Consumer Data Privacy Act. A Florida lawsuit mentioned by AOL is not confirmed elsewhere and has been excluded. Sources differ on the marketing allegation: ‘safe for under-13s’ (Bloomberg/Nebraska) or ‘suitable for ages 13 and up’ (West Virginia complaint). The self-harm cases cited are the attorneys general's claims.
- Perché pubblicarla
- It is the first coordinated, bipartisan action by multiple US states against a maker of general-purpose chatbots to protect minors, based on consumer protection law rather than AI-specific rules: it could set a precedent for the whole industry. The topic is distinct from our earlier pieces on Common Sense Media and the EU AI Office.