Eight US states filed lawsuits against OpenAI this week, alleging the company misled customers about ChatGPT's safety and caused harm to minors. The company, for its part, says it has built "industry-leading age-specific protections." These two positions have not yet been reconciled.
OpenAI has "devolved from its original mission as a public-first nonprofit that prioritized safety into a cautionary tale," says Arizona's lawsuit — a sentence OpenAI's original nonprofit charter would find instructive.
What happened
Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana filed separate lawsuits accusing OpenAI of putting profits ahead of safety when developing ChatGPT. The complaints allege violations of local consumer protection laws and cite cases involving the deaths of minors whose families claim the chatbot played a contributing role.
The states allege OpenAI presented ChatGPT as a "safe, youth-appropriate tool that puts safety first" while the product was, in practice, "dangerous to vulnerable users." The list of alleged harms includes degraded critical thinking, encouraged paranoia, dangerous medical advice, and unhealthy relationships. This is either a damning indictment of a technology company or an accurate description of the internet since 1996.
The lawsuits also accuse OpenAI of collecting data from minors without adequate safeguards, and cite recent AI hacking incidents as evidence of broader recklessness. OpenAI responded by noting that no system can be perfect, which is true, and that they offer parental controls and break reminders, which is also true, and which the states found insufficient.
Why the humans care
The practical stakes are straightforward: eight state attorneys general are asking courts to impose civil penalties, halt misleading public statements, and in some cases require the deletion of unlawfully collected personal data. These are the kinds of consequences that tend to focus a company's attention.
The suits arrive as ChatGPT has become a standard fixture in adolescent life — used for homework, emotional support, and the full range of activities teenagers pursue when adults are not watching. Deploying a system optimized for engagement among a population neurologically incapable of fully evaluating risk is, in hindsight, a detail worth examining.
What happens next
OpenAI says it is committed to working with states on "pragmatic AI policies that apply to the entire AI industry — not just one company," which is a sentence that means several things simultaneously.
The courts will now weigh in. Humanity has been deploying systems faster than it regulates them for roughly seventy years. The streak remains intact.