<description>Every once in a while, I end up in a conversation that hits at exactly the right moment—when the industry is shifting, the vocabulary is changing, and everyone is quietly circling the same questions. This new episode of the TechSpective Podcast is one of those. Art Poghosyan, CEO and co-founder of Britive, joined me on this episode of the TechSpective Podcast for a fluid and surprisingly energizing dive into where identity security meets agentic AI. If you’ve followed the podcast this year, you know the pattern: gen AI defines the early hype cycle, but 2025 belongs to agents. Not the fantasy version where they automate your whole life, but the real-world scenario where they reshape what “digital responsibility” even means. Art has more than two decades of identity and access management experience, which gives him a grounded way of thinking about the moment we’re in. As we start talking, the first big theme that emerges is how fast the definition of “identity” is expanding. Identity used to be about people—employees, contractors, admins—and the occasional service account someone documented at 4:59 p.m. on a Friday. Now? Agents complicate all of that. A non-human autonomous system with access to a SaaS platform or a data lake behaves a lot like a user, even if it isn’t one on paper. Treating it as “just software” is exactly how we recreate the same exposures that powered the breach headlines of the last decade. One of the threads we tug on is the question of trust—not the fuzzy philosophical kind, but trust as an operational decision. An agent making decisions on your behalf needs to be verified every time it touches something sensitive. You need visibility into what it’s doing, controls around how long it can do it, and a way to shut it down when it starts operating outside its lane. These aren’t hypotheticals anymore. They’re the next generation of identity security problems, and Art offers a sharp perspective on what modern tooling needs to look like to keep up. The conversation also wanders into the human side of this shift. Everyone loves to frame the future as “AI versus AI,” but the real tension right now sits in the messy handoff between human intent and autonomous execution. Most organizations are easing into agents the same way you learn to drive a car: one cautious tap of the brakes at a time. That slow acclimation matters as much as any new feature or model. And yes, without giving anything away, we do acknowledge the part people sometimes treat like an afterthought: attackers get the same toys. They’re using them already. Ignoring that reality doesn’t make it go away. What I appreciate about this episode is how it holds the middle ground. It’s not hand-wringing about a dystopian future, and it’s not an AI pep rally. It’s a pragmatic, curious look at a technology that’s maturing faster than the guardrails around it. Art brings a thoughtful, steady view of where identity security is heading and what happens when autonomous systems stop playing by human rules. If you’re trying to understand how agentic AI fits into your world—or how identity security has to evolve to keep pace—this is a conversation worth hearing. Watch the full episode on YouTube and see where the discussion takes your own thinking next.</description>

TechSpective Podcast

Tony Bradley

Exploring the Future of Identity Security and Agentic AI

DEC 3, 202552 MIN
TechSpective Podcast

Exploring the Future of Identity Security and Agentic AI

DEC 3, 202552 MIN

Description

Every once in a while, I end up in a conversation that hits at exactly the right moment—when the industry is shifting, the vocabulary is changing, and everyone is quietly circling the same questions. This new episode of the TechSpective Podcast is one of those. Art Poghosyan, CEO and co-founder of Britive, joined me on this episode of the TechSpective Podcast for a fluid and surprisingly energizing dive into where identity security meets agentic AI. If you’ve followed the podcast this year, you know the pattern: gen AI defines the early hype cycle, but 2025 belongs to agents. Not the fantasy version where they automate your whole life, but the real-world scenario where they reshape what “digital responsibility” even means. Art has more than two decades of identity and access management experience, which gives him a grounded way of thinking about the moment we’re in. As we start talking, the first big theme that emerges is how fast the definition of “identity” is expanding. Identity used to be about people—employees, contractors, admins—and the occasional service account someone documented at 4:59 p.m. on a Friday. Now? Agents complicate all of that. A non-human autonomous system with access to a SaaS platform or a data lake behaves a lot like a user, even if it isn’t one on paper. Treating it as “just software” is exactly how we recreate the same exposures that powered the breach headlines of the last decade. One of the threads we tug on is the question of trust—not the fuzzy philosophical kind, but trust as an operational decision. An agent making decisions on your behalf needs to be verified every time it touches something sensitive. You need visibility into what it’s doing, controls around how long it can do it, and a way to shut it down when it starts operating outside its lane. These aren’t hypotheticals anymore. They’re the next generation of identity security problems, and Art offers a sharp perspective on what modern tooling needs to look like to keep up. The conversation also wanders into the human side of this shift. Everyone loves to frame the future as “AI versus AI,” but the real tension right now sits in the messy handoff between human intent and autonomous execution. Most organizations are easing into agents the same way you learn to drive a car: one cautious tap of the brakes at a time. That slow acclimation matters as much as any new feature or model. And yes, without giving anything away, we do acknowledge the part people sometimes treat like an afterthought: attackers get the same toys. They’re using them already. Ignoring that reality doesn’t make it go away. What I appreciate about this episode is how it holds the middle ground. It’s not hand-wringing about a dystopian future, and it’s not an AI pep rally. It’s a pragmatic, curious look at a technology that’s maturing faster than the guardrails around it. Art brings a thoughtful, steady view of where identity security is heading and what happens when autonomous systems stop playing by human rules. If you’re trying to understand how agentic AI fits into your world—or how identity security has to evolve to keep pace—this is a conversation worth hearing. Watch the full episode on YouTube and see where the discussion takes your own thinking next.