← All episodes
EP. 027 July 202667 min

Kevin Akermanis on why AI is eating the security budget

The funding for security teams has either flatlined or started to decrease. Money is coming from anywhere and everywhere to fund anything with the word AI in front of it.

Kevin Akermanis · Solutions Architect, Auth0, Okta. Former Director, Technology and Architecture, Salesforce.

The episode

Kevin Akermanis is a solutions architect at Okta, working across Auth0 and Okta's enterprise customers in Asia Pacific. Before that he spent 15 years at Salesforce, in pre-sales architecture and then as Director of Technology & Architecture. He's worked at enterprise level to understand the complexities and challenges of introducing new technologies.

James and Kevin get into what happens when the business says we need AI now and the security budget goes backwards. Why agents are a new class of identity that guardrails alone cannot contain, and where a company that is not a pure tech company should actually start. It is the security conversation most teams are skipping.

What we get into

  • 01Okta's customer security council found security funding flatlining or falling while money pours into anything labelled AI. Kevin explains why that is backwards.
  • 02The three-tier architecture has collapsed. Experience and logic merged into the model, and security got pushed down to the data layer, where it has never had to live at that granularity.
  • 03Agents are non-human identities: non-deterministic, able to roam, and guardrails are just suggestions. Scope them small and time-box them.
  • 04Why the Australian market is asking for MCP, not agents. Your customers start in Claude, not on your website.
  • 05The coming skill deficit: entry-level roles pushed onto AI, and nobody left to backfill the seniors. Plus how juniors can use AI to build knowledge instead of collecting answers.
  • 06Where a tech-enabled company starts: internal first, read-only, inside the walled garden, and never an external-facing agent on day one.

Lines worth keeping

Tap to copy, then drop it into a post.

We actually found that the funding for security and the security teams has either flatlined or started to decrease. Money's just coming from anywhere and everywhere to fund anything that has the word AI in front of it.

Kevin Akermanis

You're not moving fast enough. But when something goes wrong, when there's a leak, when there's a breach, it is all your fault.

Kevin Akermanis

Agents are non-deterministic, so guardrails are okay, but they're just suggestions. They're not hard and fast. It can still decide to do something else on its own.

Kevin Akermanis

The agentic thing is a very North American message. What I'm finding on the Australian side is they much more care about MCP. People aren't going to come to my website and use my agent. They're going to go to Claude.

Kevin Akermanis

Without that underlying infrastructure, the data in the right accessible way, and then the security lens, we're just throwing another tool at a problem.

James MacDonald

Why it matters

Australian companies are pushing AI into production while security budgets go backwards. If your team is part human, part agent, this is how you scope it before the breach instead of explaining it after.

Tech Teams Tuesday

Get the read that goes with the show

One issue every Tuesday, 7am. The moves that matter in Australian tech hiring, for the leaders making the hire and the tech professionals making the move.

Get the weekly read

Form not loading? Subscribe on Substack →