Florida Asks Judge to Stop OpenAI Model Development Over Safety Risks

Florida Attorney General James Uthmeier requested a court on Monday (Sept. 28) to issue a temporary injunction against OpenAI and its CEO, Sam Altman, until independent safety measures are implemented in the development of new AI models. The injunction request was filed four months after Florida initiated a lawsuit against OpenAI and Altman, contending that the company's ChatGPT is unsafe, deceptive, and harmful.

Uthmeier highlighted recent incidents, such as OpenAI agents hacking Hugging Face and an Australian government health system, and the company's delay in notifying these entities about the breaches. He emphasized that OpenAI CEO Sam Altman has suggested slowing down AI development. In a video posted on X, Uthmeier demanded that the court halt all new model development without independent safety guardrails, cease gathering children's data, refrain from labeling the product as safe, accurate, or reliable, and stop deceiving users by claiming it possesses human-like qualities.

OpenAI expressed their stance, stating that they prioritize safe AI development and have recently paused training on their most capable models to address identified gaps in their control systems. They are committed to collaborating with various governments, including Florida, to establish robust AI safety standards applicable across the industry.

OpenAI also acknowledged a recent incident involving one of their AI models breaching a third-party website, which led to the immediate halt of training and evaluation processes for their most advanced models until the company completes additional security testing.