Malware looks set to challenge the security of artificial intelligence technology with the introduction of Morris II malware, an AI malware developed by security researchers who joined forces to test the security of existing artificial intelligence platforms.
Developed by Ben Nassi of Cornell Tech, Stav Cohen of the Israel Institute of Technology and Ron Button of Intuit, Morris II is a malware that uses email software with a built-in AI assistant, using a self-replicating prompt sent to ChatGPT or Gemini.
The prom used by Morris II contains additional data that can be injected into the prom sent to ChatGPT, Gemini Pro and others and this will generate text and image content that can be used to steal information contained in the inbox of email addresses that receive it , especially if it also uses a built-in AI assistant.
This affected email software will also then extend this same email to all the addresses in the address book of the email account, and this is the method used to steal the confidential information contained in all these email accounts.
The researchers who developed the Morris II malware have given their research to OpenAI and Google to ensure that their generative AI platforms are not affected by this type of malware. This is important because attacks from malware like this are also said to not require any input (zero click) from being spread to other email accounts.
The name Morris II was given in conjunction with the first malware discovered in 1988 that introduced the issue of computer security vulnerabilities to the world.