What’s the Biggest Challenge in Implementing Data & AI Governance?

I recently ran two LinkedIn polls to compare the challenges organizations face in implementing data vs. AI governance. The results were both insightful and surprising.

Priority Gap

52 percent of respondents said low organizational priority is the biggest challenge for data governance — but only 18 percent said the same for AI governance.

This is striking, considering AI governance relies heavily on strong data governance foundations. AI systems are data-driven: data goes in, models process it, and data comes out. Many AI capabilities are enabled by existing data and metadata practices. I explore these dependencies in my new book Aligning Data and AI Governance, now on Amazon.

Expertise and Resources

55 percent identified limited expertise and resources as the top challenge for AI governance, while 38% said the same for data governance.

The higher percentage for AI is understandable—it’s a newer discipline. But the figure for data governance is still high, despite decades of industry focus. I believe this correlates with the low priority given to data governance, which limits investment in upskilling.

Personally, I started my own journey in data governance from scratch, building on my finance and ERP background. Upskilling internal staff remains, in my view, the most effective path forward. For those pursuing that route, I recommend the 2025 edition of my book The Data Management Toolkit 2.0.

Regulatory Clarity

Unclear regulations mainly affect AI governance. Global AI legislation varies widely — some binding, others voluntary, some risk-focused, others principle-based. I recently reviewed multiple regulations and found over 40 unique governance principles. No wonder many organizations struggle to navigate them.

Framework Maturity

Compared to other challenges, immature frameworks are less frequently mentioned. In data governance, we have mature models like DAMA-DMBOK2 and DCAM, though implementation remains a hurdle. For AI governance, existing frameworks are fragmented and not yet aligned.

📘 For those looking to make sense of this evolving landscape, I dive deeper into these topics in Aligning Data and AI Governance, now available on Amazon.