Today, I’m excited to welcome Anshuman Bhartiya, an AppSec tech lead at Lyft. Before that, he worked as a security engineer at companies like Thirty Madison, Intuit, and Atlassian.

Anshuman is also a fellow podcaster and co-host of the Boring AppSec podcast, alongside one of my previous guests, Sandesh Mysore Anand.

Recently, he’s been experimenting extensively with building AI agents for both offensive and defensive security, and he’s documenting his findings at anshumanbhartiya.com(link in the description).

In this episode, we dive into the challenges of building effective AI agents, the impact of AI on security practices, and the importance of understanding AI outputs and avoiding confirmation bias.

We also touch on the ongoing debate of build versus buy solutions and explore where the future of AI in security might be headed.

Dive right in!