Josh Davies spoke to Diana Blass at SixFiveMedia about the security risks that should be top of mind when using generative AI applications such as ChatGPT.
Josh’s segment can be found at 13:14
Originally published in The Six Five Podcast.
Excerpt:
“Attackers can get into OpenAI’s environment and sit there quietly. If you have millions of people using this across the world, that is the perfect candidate for a supply chain compromise. If threat actors compromise OpenAI, they can use legitimate or trusted delivery methods to push out other compromises and maybe get people to download some malware. They can start giving people a code that has vulnerabilities or back doors for the threat actor organization to come and stroll in. I can see how that can be accomplished with something like ChatGPT.”
Watch the full podcast here.