%20(1).png)
We can’t build enough AI agents to keep up with cybersecurity
AI can write code, uncover vulnerabilities, and keep working long after a human logs off. Yet Ashish Rajan is not convinced security teams are about to run out of work. If anything, the opposite may be true. In this episode of Risk Grustlers, Ashish Rajan (CISO at TechRiot.io and author of AI Security Engineering) joins Nicholas Muy (CISO and VP of Engineering at Scrut Automation) to talk about what happens when AI adoption moves faster than the security practices meant to contain it. They look past the familiar question of whether AI will replace security jobs. The harder question is what it takes to turn raw model power into something defenders can rely on when an incident is unfolding, organizational context matters, and every new capability creates more to secure.
Listen on Your favourite platforms
Description
Give a security team one of the world’s most capable models, and you have not necessarily given them a useful security tool. It still needs the right context, engineering, access, and a way to act safely inside that organization.
Ashish and Nick unpack what sits between model access and practical value, including agent harnesses, incident data, model restrictions, human judgment, and the tradeoffs between closed frontier systems and open-weight alternatives. They also ask whether AI is reducing security work or simply changing where that work piles up.
For security leaders and practitioners, this episode offers a grounded view of where AI can help defenders today, where the tooling still falls short, and why security careers may become more demanding rather than less relevant.
What listeners will take away:
- Why access to a powerful model is only the starting point, not a working security capability
- How agent harnesses, organizational context, model restrictions, and open-weight alternatives shape what defenders can do in practice
- What AI could change about incident response and security careers, and why more automation may still create more work
Quote from the Episode
“Having access to a powerful AI model doesn’t mean you know how to use it well. It’s like having a Ferrari in your driveway and sitting there with the key, unable to get it out of first gear. You still need real engineers, proper harnesses, and the context of your own organization to make it useful.”
— Ashish Rajan, CISO, TechRiot.io
About the Risk Grustlers Podcast
Risk Grustlers is a podcast for people working in security, risk, and governance who want sharper conversations than the usual industry soundbites. Each episode features CISOs, security leaders, and risk practitioners sharing how they think through the real operational challenges behind cybersecurity, compliance, AI governance, and enterprise risk.
Hosted by Aayush Ghosh Choudhury (CEO and Co-founder of Scrut Automation) and Nicholas Muy (CISO at Scrut Automation), this podcast series focuses on practical lessons, hard-earned perspectives, and the nuance that comes only from years spent in the security and compliance space.
%201%20(1).avif){kind=icon}
Related Podcasts
