Researcher tricks ChatGPT into revealing security keys – by saying “I give up”

Researchers reveal how attackers can exploit vulnerabilities in AI chatbots, like ChatGPT, to obtain malicious information.