AI leadership decisions in regulated environments - less hype, more criteria.
I’m Andreas Lezgus. Four decades in public-sector IT, the last years as CTO of a large federal security organization in Germany. I write for the people who have to sign off on AI decisions: CIOs, CTOs, enterprise architects, and the executives who answer for what AI does in production.
Every issue rests on three convictions: accountability needs a name, sovereignty is the capability to switch, and context beats model. The full argument is here:
What I believe about AI in regulated environments
New here? Three pieces readers tend to start with:
Agentic AI: The flight to the wrong Portland - why autonomy is a property of context, not of the model.
Sovereign AI is not a real estate question - what a shutdown day teaches about the capability to switch.
Your AI governance has a name problem - why a register with named owners beats another committee.
I publish every other week. What you won’t find: courses, vendor promotions, or the pretense that any of this is simple.
If something resonates — or provokes — hit reply. I read every response.
Andreas
Find me on LinkedIn: https://www.linkedin.com/in/andreaslezgus/
Companion blog in german: https://www.lezgus.de
Posts reflect my personal views. They do not speak for my employer.
