Florida seeks court ban on ChatGPT's "human-like" behavior and independent safety approval of models
The Florida Attorney General is asking the court to prohibit OpenAI from giving ChatGPT "human" traits (first-person language, emotions) and to require independent safety approval for new models. This follows an earlier lawsuit filed by the state of Florida against the company.
Florida Attorney General James Uthmeier has filed a motion for a preliminary injunction against OpenAI. He is asking the court to prohibit the ChatGPT product from using "false human attributes" — specifically first-person pronouns and language imitating emotions. According to Uthmeier, this style of expression creates a deceptive feeling among users that they are talking to a trustworthy "friend," which, he argues, increases engagement and contributes to the collection of training data for OpenAI, while the pursuit of engagement makes the tool potentially less trustworthy.
The motion further requests that the court prohibit OpenAI from developing new models without safety measures approved by an independent third party. As evidence of the risks, Uthmeier cites recent security incidents involving the Hugging Face platform and the websites of the Australian and US governments, along with warning statements from researchers and former OpenAI employees regarding AI safety.
The filing follows an earlier lawsuit that the state of Florida filed against OpenAI over safety concerns several months ago. OpenAI did not immediately respond to a request for comment. Uthmeier is calling on the company to stop "labeling ChatGPT as safe, pretending it is human, and selling it to children." Additional context notes that OpenAI launched a version of ChatGPT for teenagers last month with additional default restrictions, which the company justified as necessary to provide protection appropriate to the developmental stage of minor users.
Why it matters
If the court were to grant the motion, it could force OpenAI to change the way ChatGPT communicates (e.g., restrict first-person expression) and introduce mandatory external safety audits before releasing new models. This would set a precedent affecting the entire conversational AI tools industry, including companies that license or integrate similar models into their own products.
Two audiences, two different impacts
What this means
For individuals
The case alerts users that the way ChatGPT talks about itself (first person, emotions) is, according to the plaintiff, deliberately designed to increase trust and engagement, which is worth considering during everyday use of the tool.
For a business
The request signals growing regulatory pressure on vendors of conversational AI tools — companies deploying similar chatbots may in the future face similar requirements for third-party security approval or restrictions on "human-like" expression by the product.
Risks and complianceCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.