Arrested DevOps · Matt Stratton, Trevor Hess, Jessica Kerr, and Bridget Kromhout

How AI Is Changing the SDLC With Hannah Foxwell and Robert Werner

October 1, 2025·40 min·2 clips

Matty Stratton asks if we can trust AI agents now that they're entering the old silos between testers, developers, and ops.

AI does not magically erase silos. The cold open starts with an old DevOps sore spot: testers doubting developers, ops doubting testers, and now teams trying to decide what kind of trust AI agents have earned. Matty frames AI adoption as a culture problem as much as a tooling problem. Tools matter, but incentives decide what people do with them. Trust, but verify becomes a working habit here, not poster copy. The management piece shows up early, especially through Robert's thread on trust and management change. Matty uses that to tie AI back to the same patterns DevOps has been wrestling with for years. The joke is dry, naturally. DevSecOps gets name-checked with some anxiety, mostly as a jab at how often the industry renames work it should already be doing. Then the books detour gets surprisingly useful. Matty says the most important DevOps book is not Continuous Delivery or Accelerate, but Freakonomics, because incentives explain how organizations actually move. That point fits AI cleanly. Reward the wrong behavior, and the AI rollout will faithfully amplify the wrong behavior. The episode does not pretend culture can be commanded into shape. It treats culture as something that changes when incentives and habits change. The conversation keeps hovering between blanket distrust of AI and the fantasy that agents are teammates with perfect judgment. No neat manifesto, which is probably for the best. The practical lesson stays intact: the system includes people, incentives, and verification loops. The close points listeners to the AISDLC show notes, Hannah's online community work, and resources from Hannah and Robert.

As heard by us

A grounded SDLC conversation about AI agents, trust, verification, and the incentives teams create around them.

AI in the software delivery lifecycle lands here as a people problem before it is a tooling problem. Matty Stratton frames the conversation around trust, picking up the old DevOps tensions between developers, testers, and ops and asking what happens when AI agents join that…

Read the full review in PlayNext →

Why you'd press play

If you are trying to use AI agents in software delivery without losing trust, start here.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Arrested DevOps