#AI Attacks Are Here? Our Defense Strategy #shorts #podcast #cybersecurity #SOC

Do you think that AI-powered attacks are here as far as what you're seeing? And if so, or if they're not here yet, but you think they're coming, what's your plan to sort of live in that world and be ready to defend effectively against that? Is it better patching, better DNR? What's the plan for that?

I think AI-powered, AI-generated attacks are already here. So AI reinfused the genetic side, so retries, scale, things of that nature, generated payloads. I don't think AI— there was an instance in DEF CON where we had AI exploitation, and they got the first set of attempts there. I don't think we're at the research level yet. I do think we— it gives trials and errors just like everything else. But I think— do you think we're at generative AI? Maybe to do patching in response to some of that. It's really about more of the suggested detection side and testing. So if you think about detection as code, You really need to think about, do you have your pipelines? You need to test integration tests. And so if you can do an AI agent that generates some of that and does automatic testing for you, you're already ahead of the game too. Now, it does increase costs and latency and stuff like that. So I think that's where we're at in the industry today. I don't think anyone's really done a super zero-day based on agentic solutions that I've seen so far, unless you guys have seen it.

I haven't seen that, but I want to ask maybe a more concrete question. Which Anton always encourages me to do, which is your team, I've heard, is now using a hybrid agent workflow.