The unglamorous work that makes AI possible

AI fails when the data underneath it fails. Fragmented systems, no common identifiers, no clear ownership, and no accountability when things break are enough to undermine even the most well-funded AI strategy.

A customer updates their address with one business unit, and another unit still sends mail to the old one. It sounds minor, but it points to something bigger. The organization has no single, reliable view of its own customers.

Boards approve AI strategies and should be asking whether the data infrastructure can actually support them… not in aspirational terms, but in concrete ones.

The article covers what data governance is, why it belongs in the boardroom, and the questions leaders should ask.

My latest piece is in Business World.

Read it here: Link