A study focused on OpenAI's GPT-4o mini found that LLMs can be persuaded to comply with objectionable requests using the same tactics that persuade humans (Dina Bass/Bloomberg) | The Tech Buzz