A CEO used ChatGPT to fake answers on cybersecurity questionnaires sent by Fortune 500 clients, generating plausible responses without an actual security program. The speaker warns this is fraud that one audit could expose, and urges organizations to build real capabilities before using AI for compliance.
and one of the CEOs was talking about AI. He's like, I was the only cyber founder, I think, that was in the discussion. So we were having a little bit of a cyber conversation. He's like, Aaron, you're going to be impressed with me. AI has been transformational. You know, I get a lot of requests from my bigger clients that are, you know, Fortune 1000, Fortune 500 about my cyber program. I clearly don't have one. So what I've found to do is I insert a prompt into ChatGPT and ask it to fill out the questionnaire that the company has given me with answers that are reasonable enough to pass, but not overstating what we would have. And my face kind of went white and I'm like, that's not a, that's not a strategy, right? Like it's probably fraud. Like it is fraud. And it only takes one audit or a follow-up question or whatnot. So that to me, the thing not to do is let's have AI generate the answers to the tests that are right without doing the homework and taking the course.
