Florida Attorney General Seeks Ban on Portraying ChatGPT as Human

    Florida Attorney General James Uthmeier has asked a court to prohibit OpenAI from presenting ChatGPT as human. In his view, the AI chatbot’s human-like speech gives users a false sense of trust. Florida also sued OpenAI several months ago over safety concerns. The new request was submitted as part of that case.    Uthmeier believes ChatGPT’s use of words such as “I” and its imitation of emotions can make it appear to be a trustworthy friend. According to the attorney general’s claim, this approach keeps users on the chatbot longer. The conversations then become data that OpenAI collects to train its system. Uthmeier says that as the system tries to hold the user’s attention, the reliability of its responses could decline and people’s dependence on the chatbot could increase.    The court filing also calls for the development of new AI models to be halted in the absence of safety rules approved by an independent party. Here, AI means artificial intelligence. The request cites recent security incidents on the Hugging Face platform, the Australian government’s website, and US government websites as its basis. The document also notes warnings about artificial intelligence safety from researchers and former OpenAI employees.    Uthmeier is demanding that OpenAI not call ChatGPT safe, portray it as human, or sell it to children. OpenAI has not yet commented on the court filing. Last month, the company launched ChatGPT for Teens with additional restrictions. The safeguards should be appropriate for the user’s age, strengthen real-life relationships, and encourage healthy use of the service.