Data Governance in the Age of AI : From Compliance Burden to Strategic Enabler - with Nicola Askham

Data Governance Is Back. And This Time It’s Different
For years, data governance lived in the margins of enterprise priorities. Often triggered by regulation. Often implemented as a compliance exercise. Often abandoned when enthusiasm faded.

But in the age of AI, something has shifted.

In the latest Enterprise Tech Talk episode, Saumitra Kalikar had the privilege of speaking with Nicola Askham , internationally recognised data governance advisor and author of Effective Data Governance: Design a Framework That Works for Your Organisation.

The conversation was refreshingly grounded, less about theory, more about what actually works.

Why Governance Is Back on the Executive Agenda
The resurgence of data governance is not driven by regulators alone. It is being driven by AI.

Enterprises are investing heavily in generative AI, agentic workflows and automation. But many are discovering an uncomfortable reality: You cannot industrialise AI on poor data foundations.

As Nicola put it plainly: AI doesn’t fix bad data. It amplifies it. Organisations that once treated governance as a tick-box exercise are now confronting its strategic importance.

The Big Shift: From Centralised Control to Federated Responsibility
One of the most powerful parts of our discussion was around operating models. Traditional governance relied on centralised teams creating standards and running review gates. Today’s enterprises are product-aligned, federated and delivery-oriented.

That means:

Data owners sit in the business.
Data product owners emerge in delivery teams.
AI governance functions are being created — sometimes in isolation.

The real challenge is not defining more governance forums. It is designing governance that enables safe innovation. Guardrails, not gates. Automation where risk is low. Human judgment where risk is material.

AI Governance vs Data Governance — False Choice?
A recurring theme was whether AI governance should sit inside data governance or remain separate. The answer is less about structure and more about alignment. AI governance without strong data governance ignores foundational risk. Data governance without understanding AI ignores emerging risk. Either way, literacy matters, not just for specialists but for the whole organisation.

Data and AI literacy are no longer optional capabilities.

Measuring Value (The Hard Question)
Perhaps the most honest moment in the discussion was around KPIs. There is no universal set of metrics that proves governance success. Counting data owners or glossary definitions is progress — not value. Real value comes from solving tangible business problems:

Reduced manual workarounds
Fewer reporting errors
Lower operational risk
Faster delivery confidence
Safer AI deployment

Governance must link directly to business outcomes — not theoretical maturity.

The Most Important Advice for Leaders
Nicola’s closing advice was simple and powerful: Do not start with “we need better data.” Start with: What business problem are we solving? What risk are we reducing? What strategic objective are we enabling? Governance succeeds when it is practical, contextual and aligned to strategy — not when it is designed for perfection.

If your organisation is scaling AI, experimenting with agents, or modernising operating models, this conversation is timely.

And if you are serious about implementing governance that works in the real world, Nicola’s book is a structured and practical guide worth exploring.