#126 - ChatGPT & Generative AI (with Konstantinos Sgantzos)
CISO Tradecraft®
00:00
How to Armorize Generative AI
Security leaders need to be asking, how do we put safeguards in their generative AI like a web app firewall? If those safeguards are not designed by the writers of the GAI tools, then we'll have to add third party software for protection. There is a team of people in OpenAI or Microsoft or Google that are actively preventing several bad questions from being answered. I believe that this will probably form a database that contains everything that needs to be so you cannot ask these questions in any way. A proper definition of what is considered illegal needs to be in place. It includes an identity for every AI agent out there and anonymity for the user until proven to commit a malicious act.
Transcript
Play full episode
