The AI company’s investigative team found that many accounts were using the program to engage in malicious activity around the world, such as employment schemes, social engineering, and cyber espionage.