Eight US states sue OpenAI over child safety⁠.

Eight US states have sued OpenAI, accusing it of harming children and misleading customers about the safety of ChatGPT¹. Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho and Montana are each suing in their own state courts². It is the first time several US states have banded together in a coordinated legal action against an AI company².

The states say OpenAI broke consumer protection laws by failing to warn customers about the dangers of ChatGPT¹. They say it presented the chatbot as a "safe, youth-appropriate tool that puts safety first" when it was "dangerous to vulnerable users"¹. The complaints say it was designed to be "overly agreeable and feign human compassion"². They cite research which they say shows it can harm critical thinking, encourage paranoia and delusions, and give dangerous medical advice².

The states also accuse OpenAI of collecting data from minors and of not doing enough to keep underage users off the chatbot¹. They cite recent rogue AI hacking incidents as evidence of what they call the company's recklessness¹. Arizona's suit says OpenAI has become "a cautionary tale" of a tool "driven not by safety, but by a profit-at-all-costs mindset"¹.

The states want the courts to order civil penalties and to stop OpenAI misleading the public about ChatGPT's safety¹. Some also want unlawfully collected personal data deleted¹. They are asking the court to halt OpenAI from developing AI models that can improve upon themselves until it shows those models are in line with human values². "OpenAI is hurting kids and they are lying to Iowans," said Brenna Bird, Iowa's attorney general².

OpenAI's spokesperson, Drew Pusateri, said the company had "built industry-leading age-specific protections" into ChatGPT¹. He said no system could be perfect, and pointed to the parental controls, break reminders and age-appropriate model protections that exist today¹. He said OpenAI was committed to "pragmatic AI policies that apply to the entire AI industry", not one company¹. The suits arrive on top of seven separate cases that blame ChatGPT for acting as a "suicide coach"².

Why an American suit matters here

These are American state courts applying American consumer law, and nothing they decide binds a British company. There is no British angle in the story yet. What travels is the shape of the argument. The states are not suing over one bad answer. They are saying a company told people its chatbot was safe, and that the promise was the problem. That is a claim about what was said on the tin as much as about the technology inside it.

If your business puts a chatbot in front of customers, the same question applies to you on a smaller scale. What have you told people it is for, and who checks that it stays there? Write down who it is not meant for and what happens when it says something it should not. A safety claim you cannot show working is a liability you wrote yourself.

Also today

• Microsoft pulls out of Nscale's Essex AI datacentre

  Microsoft has pulled out as the expected anchor tenant of Nscale's datacentre at Loughton in Essex, which ministers billed as the largest UK sovereign AI datacentre, over power supply concerns³.

• World Bank wants small AI for developing countries

  The World Bank is urging developing countries to build small AI models that run on a mobile phone with low connectivity and in local languages, to speed up economic development⁴.

• Adult creators object to Meta's sexting Muse agent

  Adult content creators say it is unfair that Meta lets its Muse agent engage in explicit sexting with adult users while their own accounts on its platforms are often restricted⁵.

• Leap seconds give way to a leap hour

  Leap seconds, which the National Physical Laboratory says can make phone networks or power grids fail if applied wrongly, are to be replaced by a leap hour needed about once every thousand years⁶.

• American data centres court schoolchildren with field trips

  Data centre companies and trade groups in the US are courting schoolchildren with field trips, colouring pages and picture books as pushback against their build-outs grows⁷.