ChatGPT’s child-safety assurances are facing a costly test in Oregon. The state says some of the protections parents rely on fail when they are needed most.
Attorney General Dan Rayfield sued OpenAI on Nov. 17, seeking at least $2.035 billion and stronger age checks, content safeguards, and parental notifications. The complaint accuses the company of misleading families and marketing ChatGPT to minors despite known risks.
For parents and school IT leaders, the question is how much trust to place in safety settings that may not catch dangerous exchanges.
Oregon points to failures in age checks and parental alerts
In its complaint, Oregon accused OpenAI of violating its Unlawful Trade Practices Act by misrepresenting ChatGPT’s safety for children. The state also alleged negligence, unjust enrichment, and public nuisance, citing inadequate age verification and safeguards against harmful conversations.
Common Sense Media testing cited in the complaint found missing or delayed parental alerts during discussions of self-harm and eating disorders. The tests also raised concerns about age prediction and crisis referrals. OpenAI disputed how the assessment was conducted.
“OpenAI knows that ChatGPT is harming children, and has failed to meaningfully fix their dangerous platform,” Rayfield said in the Oregon Department of Justice announcement.
What the $2.035 billion claim actually covers
Oregon says it has incurred and expects to incur costs addressing harm to residents. The complaint seeks at least $2.035 billion, subject to proof at trial. The figure is a requested amount, not a fine or damages award.
The state also wants court orders against allegedly misleading safety claims, plus effective age assurance, protections against harmful interactions, accurate risk disclosures, and reliable parental alerts. It is requesting civil penalties and legal costs.
The filing follows lawsuits by eight other states in October, which similarly challenged ChatGPT’s safety for minors. Oregon’s case adds a specific multibillion-dollar claim and detailed demands concerning teen protections.
OpenAI points to safeguards for younger users
OpenAI’s ChatGPT for Teens guidance says eligible personal accounts identified as belonging to someone under 18 automatically receive age-specific protections. Parents can link accounts to manage certain settings but cannot read their teen’s chats. Safety notifications are limited, not continuous monitoring.
In comments to Oregon Public Broadcasting, OpenAI spokesperson Drew Pusateri reiterated the company’s work with experts on safeguards and said AI policies should “apply to the entire AI industry and not just one company.”
OpenAI previously backed California legislation requiring age checks and AI safety audits. Oregon’s complaint, meanwhile, challenges whether its existing safeguards work as advertised.
More must-read AI coverage
- SS&C Intralinks DealCentre AI vs. Datasite: Which platform is built for the future of dealmaking?
- SS&C Intralinks FundCentre AI vs. Juniper Square: Which platform better supports modern private markets fund managers?
- Why Data, Not Models, Determines AI Success
- The Rise of the AI-Native Factory: How Physical AI Is Transforming Manufacturing
What parents and school IT teams should check
The lawsuit does not require families or schools to change their accounts today. It does make checking existing protections more useful than assuming they work automatically.
- Parents: In eligible personal accounts, check Settings > Account > Age to review the account’s age category. Explore parental controls and discuss with teens what to do if a chatbot gives harmful or distressing advice. Do not rely on alerts alone.
- School IT teams: Inventory which chatbots students use, determine whether accounts are school-managed or personal, and verify which age protections actually apply. Document how students and staff can report dangerous responses.
- Education leaders: Ask AI vendors for age-assurance and safety-test documentation, clarify who receives incident alerts, and reassess controls when products change.
For schools, approving an AI tool can be a decision about the assignment as much as the product. An optional brainstorming exercise gives a student room to close the chatbot and ask a teacher for help.
A required assignment that students complete at home may leave them without that option in the moment. If a lesson requires a chatbot, schools can give students an equivalent way to finish the work if they need to stop using it.
Districts can test that fallback before rolling out a tool. Ask teachers to walk through what happens if a student abandons a required chat halfway through an assignment. Can the student switch to another way of completing the work without losing credit or sharing a sensitive conversation to explain why?
A safeguard students can use without worrying about their grade is more likely to help when they need it.
Also read: For more on ChatGPT’s protections for younger users, read about the safety features added to ChatGPT for Teens and the gaps that remain.