tgroenwals shared this post · May 15

Governance exists to protect decisions, not data.

That line changes the whole conversation with the board. The moment you stop talking about data quality and start talking about decision quality, the room leans in.

I drew this framework after getting the same question from three different CDOs in the same month: “We have governance on our roadmap but nobody agrees on what it actually includes.”

So here are the layers, bottom to top:

Layer 1: Foundational data management. Data modelling, master data, integration, storage, security. Without foundations, every layer above is unstable.

Wayne Conrad The reframe from data quality to decision quality is the right one. Most boards do not fund what they cannot connect to outcomes they own.

The layer that collapses most often is Layer 4. Ownership gets assigned on paper during a governance workshop. Then the data steward leaves, the domain lead changes roles, and nobody updates the registry. Twelve months later the council exists but the owners do not. Policies sit in a document that nobody enforces because enforcement requires someone with authority and time. Most stewards have neither.

The AI output problem compounds this fast. Organizations are building AI on top of metrics that have no named owner at the business level. The model performs. The output gets used. The decision gets made. When the output is wrong, the accountability chain has no end point. The governance framework documented the layers. It did not survive the first reorg.

Layer 5 is where AI governance actually lives, and most frameworks treat it as consumption, not control. If the output has no owner, the decision below it has no check.

Governance without enforcement is just a diagram. May 15 2 likes

Dr. Alexander Borek A lot of governance discussions stay too abstract. Connecting it to decision quality makes it operational very quickly. May 15

tgroenwals shared this post · Apr 29

John Wernfeldt

I asked a client last month how many decisions their AI agent makes per day.

They said “maybe a few.”

We counted. It was over 600.

Which data source to query. How to interpret the metric. What context to include in the answer. Whether the output was confident enough to show a customer.

Every one of those is a decision. Every one of them had zero governance around it.

Nobody owned any of them. No documentation, no named reviewer, no exception process.

So I drew the levels of decision rights an AI agent actually has, and what humans need to own at each level.

Rajesh Ramaswami True, John. AI agents move fast until nobody owns the decisions they make. Work gets risky fast when agents trigger actions and exceptions have no owner. Apr 29 1 like

Alex Barády Important point, John. Governance gaps at the Act level can cause real risk. Apr 29 1 like

tgroenwals shared this post · Apr 17

John Wernfeldt

Every governance program needs data owners. Almost nobody writes down what a data owner actually does.

So you end up with a name on a slide and no real change.

I've been pulled into too many governance reviews where the data owner role exists in theory and disappears in practice. Nobody knows what they're accountable for. Nobody knows what they decide. Nobody knows when to call them.

Here's the job description that actually works:

Accountable for:

→ One named metric
→ Its written definition
→ Its calculation logic

Decisions they make:

→ Approves definition changes
→ Signs off…

Edris Yaghob The disconnect between accountability on paper and accountability in practice shows up everywhere, and your breakdown of specific decisions makes the difference between a RACI chart that sits in a drawer versus one that actually gets used. Apr 16 1 like

Clare Kitching Well said John, defining ownership at the metric level makes accountability real instead of just something that exists on slides. Apr 16 2 likes

Public author posts