Florida seeks a ban on ChatGPT acting like a person
TL;DR
Florida Attorney General James Uthmeier has asked a judge to stop OpenAI from giving ChatGPT false human attributes, a few months after the state sued the company over safety concerns. He argues that first-person pronouns and output that mimics emotion make users see the bot as a trustworthy friend. According to Uthmeier, that design drives engagement and feeds OpenAI more training data, while users come to rely on a system that may be less trustworthy than it appears.
Nauti's Take
Clear rules on how human a chatbot may appear would be an opportunity for more transparency, especially for kids and teenagers. The catch is that a court would have to decide when first-person language and a friendly tone already count as deception, and that line is hard to draw.
Teams building their own assistants should review how much closeness their persona suggests and whether users can clearly tell they are talking to a machine.
Sources