Recent research from Hoxhunt analyzed effectiveness of ChatGPT-generated phishing attacks showing people are still better at deceiving other people.