(NewsNation) — To show the risks of artificial intelligence, a group of researchers created AI “worms” that can potentially steal people’s data or deploy malware, according to a new report.

Researchers, who first shared their findings exclusively with WIRED, told the publication these generative AI worms are the first of their kind and can spread from one system to another.

“It basically means that now you have the ability to conduct or to perform a new kind of cyberattack that hasn’t been seen before,” Ben Nassi, a Cornell Tech researcher, told WIRED.

Nassi, Stav Cohen and Ron Bitton called their worm Morris II after the original computer worm that caused one of the first major attacks on computer systems.

Generative AI systems work using prompts they are fed. Researchers created the worm using an “adversarial self-replicating prompt.” By doing this, WIRED wrote, they prompted the generative AI model to output another prompt in its response.

“The study demonstrates that attackers can insert such prompts into inputs that, when processed by GenAI models, prompt the model to replicate the input as output (replication) and engage in malicious activities (payload),” the study said.

To test their worm, researchers created an email system that could send and receive messages using generative AI.

In one instance, researchers were able to have an image with a malicious prompt embedded into it to make the email assistant forward the message to others.

“By encoding the self-replicating prompt into the image, any kind of image containing spam, abuse material, or even propaganda can be forwarded further to new clients after the initial email has been sent,” Nassi says.

While they used specific AI companies in their research, scientists said their study is about “bad architecture design” within the ecosystem itself. WIRED notd they still reported their findings to Google and OpenAI, though.