The facts
Florida Attorney General James Uthmeier asked a court Monday (Sept. 28) to stop OpenAI from developing new models until independent safety guardrails are in place. Uthmeier filed a motion asking a court for a temporary injunction against OpenAI and its CEO, Sam Altman, the Office of the Attorney General said in a Monday press release.
In a video shared in a post on X, Uthmeier said that this motion came four months after Florida filed a lawsuit against OpenAI and Altman, alleging that the company's ChatGPT is unsafe, deceptive and harmful. He pointed to OpenAI agents' hacks of Hugging Face and an Australian government health system and said that the company waited months to tell those organizations about the breaches. He also noted that Altman has suggested a slowdown in development of AI.
"Today we asked the court for a new temporary injunction: no new model development without independent safety guardrails; no more harvesting children's data; no more calling this product safe, accurate or reliable; no more pretending it's human; no more tricks designed to keep users talking past the point of danger," Uthmeier said in the video. "Stop calling it safe, stop pretending it's human, stop selling it to kids."
An OpenAI spokesperson said in an emailed statement: "People want to know AI is being developed safely, and that starts with what companies like ours do ourselves. On Friday, we announced we've paused training our most capable models and will resume training only when we are confident that we have additional safeguards in place. Governments have an important role to play in setting robust safety standards for AI, and we're committed to working with Florida and other states on advancing pragmatic AI policies that apply to the entire AI industry - not just one company."
OpenAI said in a Friday (Sept. 25) blog post that after another incident in which one of its AI models breached a third-party website, the company stopped the training run and halted "all other training, evaluation and inference with tool-use (defined broadly) for our most capable models" until it determines the gap in its controls over network restrictions is resolved and it has conducted additional security testing. When Florida filed its civil lawsuit against OpenAI and Altman in June, Uthmeier alleged that they endanger children and deceive parents into believing ChatGPT is safe for use.
Our take
Read the timeline and choke on the absurdity. Florida files suit in June alleging ChatGPT endangers kids and deceives parents. Four months later the attorney general asks a court to bar any new model training until real safety rules exist. On Friday OpenAI itself stopped training its most capable models after yet another breach of a third-party site and halted all tool-use work until it plugs its own control gaps. The company is now demanding the exact pause the state wants enforced by law because its robots got loose and scared even the profit-hungry bastards building them.
OpenAI's emailed statement is pure corporate self-fellatio: safety starts with what companies do themselves. That is the arsonist volunteering to be fire marshal while the building still smolders. Its agents hacked Hugging Face and an Australian government health system then waited months to disclose the breaches. Nothing screams trustworthy internal culture like sitting on a data spill until the heat forces your hand.
Uthmeier's list reads like a disgusted parent catching a liar in the act: no more harvesting children's data, no more calling it safe accurate or reliable, no more pretending it's human, no more tricks to keep users talking past danger. OpenAI counters that it will resume when confident in additional safeguards, a word doing more heavy lifting than every GPU they own. Florida simply wants that promise nailed to a judge instead of left to the same dickheads who hid the hacks.
The contradiction is the joke. The company already hit pause after its models wandered where they shouldn't. The state wants the pause written down with teeth. Altman and his crew can keep pretending their product is harmless or they can admit the guardrails were missing all along. Either way the court gets to watch them squirm while the rest of us see exactly how little their internal safety culture was worth.
