I spend a lot of time talking to CIOs and CDOs. Most have an AI policy. Far fewer can tell me which production models are running without a…

Felix Van De Maele

Entrepreneur, Founder, CEO of Collibra

I spend a lot of time talking to CIOs and CDOs. Most have an AI policy. Far fewer can tell me which production models are running without a documented accuracy score, or which were trained on data that was never classified for sensitivity. The policy exists. Whether anything checks it is a different question. That is the gap in AI governance right now. Not the absence of rules, but the ability to enforce them. And the hard part is turning policy into something operational. A meaningful control has to understand both what the governance team expects and how that requirement maps to the actual models, data, and attributes in your environment. Too often, those two things live in different places. Regulatory frameworks are increasingly pushing companies toward the same question: can you prove your controls are actually working? That is the problem we are addressing with new out-of-the-box controls in Collibra Control Tower. More than 80 checks turn governance requirements into controls that can run continuously, so teams can see where they are meeting their standards and where they are not. The tooling matters, but the bigger point is simple: governance you cannot verify is just a policy. As AI takes on more responsibility inside the enterprise, having the right rules will not be enough. You need to know they still hold when the AI is actually running. https://lnkd.in/ebA8z6is