ChatGPT, Other Generative AI Apps Prone to Compromise, ManipulationBy rooter / July 25, 2023 Researchers find artificial intelligence applications that use large language models could be compromised by attackers using natural language to dupe users.