world ChatGPT mirrors abusive language in heated conversations, study finds euronews 18h ago OpenAI’s ChatGPT produced insults and swear words as it responded to tense human conflicts, a study has found. www.euronews.com