Podcast Summary: Decoder with Nilay Patel – Okta's CEO Is Betting Big on AI Agent Identity

Date: March 30, 2026
Host: Nilay Patel (The Verge)
Guest: Todd McKinnon (Co-Founder & CEO, Okta)

Overview of Episode Theme

This episode of Decoder dives deep into the seismic shifts happening in enterprise software, driven by the rapid rise of AI agents and the so-called "SaaS apocalypse." Nilay Patel speaks with Okta’s CEO, Todd McKinnon, about Okta’s strategic focus on AI, paranoia around disruption in SaaS, the emergence of agent identity as a core business need, and the wild future of org charts and software development. The conversation is both technical and philosophical, giving listeners a candid, inside look at how a security and identity company is preparing for the future and helping to define it.

Key Discussion Points & Insights

1. The "SaaS Apocalypse" and Okta’s Response

[04:49–11:28]

• Paranoia as a Strategy: Todd characterizes his own mindset (and Okta’s) as “paranoid” about the potential for disruption, especially as new AI-driven, agentic software enables both startups and companies’ internal teams to rapidly build alternatives to traditional SaaS.
  • Quote [05:33]: "We're paranoid and we're making sure that we're using all the latest technologies, LLMs, et cetera, to make sure that we have something that's resilient and secure but has the best features and best capabilities." – Todd McKinnon
• Difference in SaaS Categories: Okta’s business is somewhat insulated because identity/security software is mission-critical, must be reliable, and carries a brand promise around trust.
  • Quote [08:45]: "Even if you're building something that looks like Okta, getting the features to work is like 10% of the battle. Making sure it works 100% of the time takes years and years and years." – Todd McKinnon

2. Agent Identity: The Next Frontier

[12:46–19:35]

• New Market Emerges: Okta sees the enormous opportunity (and challenge) in providing identity and security for AI agents—a market potentially larger than human identity management.
  • Quote [15:48]: "We have the opportunity to win this battle to be the identity layer for AI agents. And if we win that, that could easily be the biggest category in cyber."
• Enterprise Realities: Organizations are well aware of the potential of agentic enterprise but face confusion due to the influx of solutions from major vendors and the lack of clear rails/standards.

3. Building Guardrails for AI Agents

[19:35–22:19, 53:53–57:02]

• OpenClaw as a “ChatGPT Moment” for Agents: The rise of platforms like OpenClaw has shown companies what's possible but also raised acute security challenges, as agents need access to core data to be effective.
• Okta’s Blueprint:
  • 1. Onboarding agents as first-class identities (not just users or machines, but something hybrid).
  • 2. Standardizing connections and permissions across siloed platforms.
  • 3. Providing a kill switch if agents go rogue.
  • Quote [54:24]: "Agents are a new identity type and it's kind of like a combination of it has some attributes of a human identity and some attributes of…a system."
  • Quote [56:53]: "Pulling the access to everything the agent can access, not access to the agent." – Todd McKinnon

4. The Growth and “Messiness” of Software Market

[22:19–25:05, 40:50–46:04]

• Software and IT Market Exploding with Agents: “The pie is getting much, much larger” for both software and IT services, with new digital worker layers emerging.
  • Quote [22:40]: "If you look at the amount spent on software...it's about $1.2 trillion. If you look at the services, the IT services market, it's about $1.8 trillion. The markets are getting bigger. We're going to be spending more of that money on software."
• Disruption and Consolidation: Some SaaS winners will churn, but security and infrastructure providers with trust and deep integration stand to persist or grow.

5. Org Charts & Decision-Making in the AI Era

[29:52–37:52]

• Radically Changing Org Charts: Okta is seeing new challenges as agents reshape the workforce. McKinnon prefers empowering strong managers and minimizing organizational shuffling but acknowledges AI is forcing change everywhere.
  • Quote [30:19]: "The guiding principle is try to give the great people area that they can be great in...It's really kind of a people driven org chart."
• Decision-Making Evolution: With company size and stakes increasing, McKinnon trusts his instincts but narrows focus on which decisions matter most, diving into details where necessary.

6. Can Today's AI Agents Bear The Weight?

[39:36–44:23]

• LLMs Limitations and Opportunity: McKinnon believes the tech will improve, but even in its current state, the transformation and the market opportunity are enormous.
  • Quote [40:50]: "Absolutely, the technology can develop. I think there's a lot of wild extrapolations going on right now, but I think that even if you don't meet the wild extrapolations people are talking about, the market is still massive."
• Future of Developers: The explosion of agentic systems will create demand for even more engineers, as new challenges emerge around maintaining, scaling, and overseeing these complex agentic workflows.

7. The Database as the Prize

[44:23–46:04]

• “Chat to the Database” Paradigm: Increasingly, value accrues to data access and control. As agents build bespoke workflows, the data layer (databases, data warehouses) becomes the most valuable commodity.
  • Quote [44:23]: "All software development in 2026 is just calibrating the interface between your brain and a database" [as quoted by a Verge designer].

8. Unbundling and the Coming Standards War

[50:23–52:02, 59:38–63:30]

• Layer Unbundling: There is a “data layer,” “intelligence layer,” and “front end layer,” which are becoming more connected and less siloed, driving both confusion and opportunity.
  • Quote [50:47]: “There’s an unbundling and there’s like a data layer and an intelligence layer and a front end layer. But what also is happening is that...all the apps that you thought were in various silos are connecting to each other.”
• Standards and Customer Power: Okta is pushing for industry standards around agent identity and access. McKinnon points out that historically, customer demand and, if needed, regulation (antitrust) have forced open markets and prevented lock-in.
  • Quote [60:36]: "There's only one thing. It's customers. I mean, customers have to have...the leverage. And if...the market mechanism [fails]...the government will step in and do antitrust."
  • Memorable moment [61:53]: McKinnon’s enthusiastic endorsement of strong antitrust if vendors don’t play fair.

9. Security, Fraud, and Digital Identity

[64:36–68:51]

• Rising Fraud Risks: As agents proliferate, so does AI-powered fraud, making robust identity all the more crucial.
• Digitizing Identity: Governments are rolling out digital national IDs, mobile driver’s licenses, etc.—tech Okta wants to support for both security and privacy.
  • Quote [65:34]: "40% of our business is authenticating and validating customers logging into customer websites and mobile apps. And this area is changing a lot with AI as well."
  • Quote [68:51]: "I do...There's technical solutions, there's also process and regulatory parts of it. I think ultimately the most privacy preserving thing is like no technology. So there's going to be a trade off..."

Notable Quotes & Moments (with Timestamps)

• Paranoia as Fuel:
  "We're paranoid and we're making sure...we have something that's resilient and secure but has the best features and best capabilities." – Todd McKinnon [05:33]

• Why Okta Survives DIY SaaS:
  "Making sure it works 100% of the time takes years and years and years. And there's also a reputational thing...Are you going to trust the proven solution that's been out there for years?" – Todd McKinnon [08:45]

• Agent Identity as a Massive Opportunity:
  "We have the opportunity to win this battle to be the identity layer for AI agents. And if we win that, that could easily be the biggest category in cyber." – Todd McKinnon [15:48]

• ChatGPT Moment for Agents:
  "It's the ChatGPT moment for agents and then ChatGPT was the Netscape moment for AI. So it's very significant." – Todd McKinnon [19:35]

• Unpacking Agent Identity:
  "Agents are a new identity type and it's kind of like a combination of...a human identity and...a system." – Todd McKinnon [54:24]

• Kill Switch for Rogue Agents:
  "Pulling the access to everything the agent can access, not access to the agent." – Todd McKinnon [56:53]

• Antitrust as Last Resort:
  "If there's so much monopolistic lock in that, that we have to rely on regulators to come in and fix it." – Todd McKinnon [61:53]

• The Future Is Not Just Boring Automation:
  "There will be another wave of things that are automating processes that weren't possible before. We're still in the early parts of that second phase..." – Todd McKinnon [53:19]

Structure & Flow of Conversation (with Timestamps)

• [01:50] Introduction by Nilay; overview of Okta, SaaS apocalypse, and agent identity
• [04:16–11:28] Todd on SaaS apocalypse, business model paranoia, and the trust/integration moat
• [12:46–19:35] The strategic pivot to agent identity and massive new cyber markets
• [19:35–22:19] OpenClaw, agent security challenges, Okta’s rails for AI
• [22:19–25:05] Expanding markets and opportunities
• [29:52–37:52] Org chart evolution, decision-making, and the changing structure of Okta
• [39:36–44:23] Can current AI tech deliver? The future of software development and developer careers
• [44:23–46:04] The importance of data and databases in the AI era
• [50:23–52:02] The unbundling of the software stack and emergence of agent standards
• [53:53–57:02] Agent onboarding, identity, and kill switches
• [59:38–63:30] Vendor lock-in, standards, antitrust, and customer leverage
• [64:36–68:51] Digital identity, fraud, age verification, and privacy challenges
• [69:37–70:02] What’s next for Okta: blueprint for secure agentic enterprises

Key Takeaways

• AI agents are fundamentally changing the structure and strategy of SaaS companies.
• Okta sees agent identity and governance as “the next frontier” in cybersecurity.
• The “SaaS apocalypse” will impact undifferentiated, non-critical tools, but infrastructure security remains robust—for now.
• Guardrails, standardization, and kill switches for agents will be vital as agent adoption grows.
• Digital identity—anchored by government-issued credentials—will be critical in fighting fraud and distinguishing humans from bots/agents.
• Revolutions in software architecture, data access, and corporate org charts are already underway.
• Industry standards and customer leverage (possibly enforced by regulation) will determine how open future agentic ecosystems become.

This episode blends technical, strategic, and even emotional perspectives on one of the most pressing issues in enterprise tech: how we secure, govern, and profit from the coming wave of AI agents and digital workers. For anyone interested in the crossroads of identity, security, and artificial intelligence, Todd McKinnon and Nilay Patel deliver a candid, insightful conversation about the challenges and opportunities ahead.