What happens when AI makes your security teams faster, but leaves the same people carrying all the risk?
In this episode of Business of Cybersecurity, I sit down with Shan Kulkarni, CEO of Nullify, to discuss a growing tension that many security leaders are already feeling. AI is helping developers ship code faster than ever. Still, for product security teams, that speed often creates even more alerts, more vulnerabilities to review, and more pressure on already stretched teams.
Shan argues that the real issue is not productivity alone. It is accountability. When copilots increase output while ownership remains with the same engineers, the workload does not disappear. It multiplies.
We explore why Shan believes the next phase of enterprise AI will be shaped by autonomous AI employees rather than assistant-style tools.
He explains how Nullify is designed to onboard, reason, and act like a human security engineer, with access to code bases, ticketing systems, cloud environments, and internal documentation.
From validating whether a vulnerability is truly exploitable to assigning fixes and following up with developers, Shan shows how AI workers could replace several disconnected security tools and the extensive manual coordination required.
Our conversation also gets into trust, which remains one of the biggest barriers to adoption in high-risk environments. Shan talks openly about the safeguards needed before companies will feel comfortable allowing AI to take action instead of simply making suggestions.
We discuss merge-ready patches, exploit confidence scores, the rising threat surface created by AI-generated code, and why authorization, authentication, and business logic flaws may become some of the biggest risks in modern software.
It is a timely conversation about what security teams actually need right now: fewer dashboards, fewer false positives, and better ways to manage growing responsibility in a world of machine-speed software delivery. If you are trying to understand where AI fits inside security operations, and whether autonomous systems can truly ease the burden rather than increase it, this episode should give you plenty to think about.
What do you think, are we heading toward a future of AI teammates in cybersecurity, and how much responsibility are you willing to hand over?
Useful Links