AI Security Podcast
Hosted by Ashish Rajan & Caleb Sima
Two former CISOs separate real AI security risk from vendor noise, for the people who have to sign off on it.
- Level
- Practitioner, assumes you work in or study security
- Status
- Active, last episode
- Who's behind it
- Independent, TechRiot.io (Ashish Rajan) sourceFeed owner and copyright: TechRiot.io, the same producer as Cloud Security Podcast.
- Last verified
- Transcripts for some episodes
Listed underAppSecCISO and leadership
Editorial take
The most useful regular show for security leaders who have to make AI decisions. Ashish Rajan (Cloud Security Podcast) and Caleb Sima, previously chief security officer at Robinhood, mix two-host discussions with guests from banks, Canva, Box, and DoorDash. Episodes run 45 to 60 minutes. The show is strongest on agent authorization, prompt injection containment, build versus buy for AI security tools, and how to explain AI risk to a board. Caleb's skepticism about AI security startups is a welcome counterweight. The limitation is depth: it is a strategy show, so engineers wanting attack walkthroughs or model internals will want something more technical.
Last hand-checked 2026-09-30.
Listen if you …
- are a CISO or security architect being asked to approve AI agents and copilots
- want a sanity check on AI security vendors before a buying decision
- already follow Cloud Security Podcast and want the AI half split out
Skip if you …
- you want hands-on red teaming or model internals, discussion stays at the strategy level
- you are new to security, episodes assume you know how enterprise security programs run
Start with these 3 episodes
- 01
· 44 min · Beginner
Plain-language grounding in how LLMs work before the show moves on to securing them.
- 02
MCP vs A2A Explained: AI Agent Communication Protocols & Security Risks
· 54 min · Practitioner
The two hosts doing what they do best: unpacking a new agent protocol and where its trust boundaries break.
- 03
The AI AuthZ Problem: Why Human Least Privilege Fails for Autonomous Agents
· 48 min · Practitioner
Gets at the identity question under every agent rollout: what least privilege means when the user is software.
About the show
AI Security Podcast is produced by TechRiot.io and co-hosted by Ashish Rajan and Caleb Sima, both former CISOs. It launched in October 2023, initially as AI Cybersecurity Podcast with a short AI Explained series on LLM basics.
Episodes alternate between host discussions of recent AI security news or conference takeaways and interviews with security leaders deploying AI. The website publishes transcripts for many episodes, and video versions are on YouTube.
Notable guests
- Phil Venables
- Kane Narraway
- Sandip Wadje
- Cezary Piekarski
Pairs with
If AI Security Podcast works for you, these likely will too.
Cloud Security Podcast
Ashish Rajan
Practitioners from Netflix, Block, and Adobe explain how they actually run cloud security, now with a heavy AI tilt.
Listen if you are a cloud or platform engineer moving into security.
interviewbiweekly30–60mUnsupervised Learning
Daniel Miessler
A security veteran's essays on where AI is taking security, work, and people, published when he has something to say.
Listen if you want frameworks for thinking about AI's effect on attack and defense.
mixedirregular30–60mResilient Cyber
Chris Hughes
Supply chain, vulnerability management and AI agent security, from a practitioner who writes the books on them.
Listen if you work in AppSec, product security or vulnerability management and want the current debates.
interviewweekly30–60mLatent Space
Pickswyx (Shawn Wang) · Alessio Fanelli
The technical AI engineering podcast, practitioners talking to practitioners.
Listen if you you ship LLM-powered features and need implementation-grade detail.
interviewweekly30–60m