The data governance program era is ending
After 20 years in data, I'd never written a post on LinkedIn. Not one. That ends today.
Here's why I'm starting now. The data governance industry has spent two decades shipping programs that nobody uses. New tool, new framework, new council. Same outcome. The people who were supposed to use it didn't. The executives who paid for it stopped funding it. And the data still got worse.
AI is about to end that era. Not by improving the programs. By making them unnecessary.
The work a data steward did in 2023 (manually classifying columns, writing definitions, chasing owners, reconciling lineage) an agent will do overnight in 2026. The accuracy gap is closing faster than most data leaders are budgeting for. I've watched it happen on real datasets.
That doesn't kill the data steward role. It transforms it. The steward becomes the AI orchestrator: the person who writes the rules an agent runs against, then watches the metric move.
That's what I'm going to write about here. What happens when AI swallows the unsexy work of data governance. What it means for the tools, the org chart, and the people who run data estates in the next decade.
I'll start at one post a week, then build from there. The plan is to be useful, not loud.
If you're building in this space, running a pilot, or you think I'm wrong, tell me. I read everything.