Researchers develop generative AI worm that can steal data and send spam emails

Researchers have developed a generative AI worm called Morris II that can steal data and send spam. The Morris II worm uses two methods: a text-based, self-replicating prompt and a self-replicating prompt embedded in an image file. In the first method, the worm uses Retrieval Augmented Generation (RAG) to "poison" an email program's database, allowing it to steal data and infect new hosts when responding to emails. In the second method, a malicious request embedded in an image triggers the email assistant to forward the message, spreading spam, offensive material, or propaganda. The researchers believe that generative AI worms could appear in the wild in the next two to three years.