ChatGPT mirrors abusive language in heated conversations, study finds

OpenAI’s ChatGPT produced insults and swear words as it responded to tense human conflicts, a study has found.