The Pre-Deployment Risk Check

Before you put an AI system in front of real people, run a pre-deployment risk check. It doesn't need to be a six-week engagement. It needs to be an honest hour with four questions.

The NIST AI Risk Management Framework, published in January 2023, organizes those questions into four plain-English functions. Govern: who's in charge of this system, and what are the rules? Map: what could go wrong, and who could it harm? Measure: have we tested it — really tested it — on the things we're worried about? Manage: once it's live, how will we watch it and what do we do when something breaks?

Turn those into a checklist and walk through it before every meaningful launch or update. Write down the answers. If you can't answer one, that's not a failure — that's the check doing its job. Fix the gap, then launch.

Teams skip this because it feels like slowing down. It's the opposite. An hour of honest questions now prevents months of painful answers later — to customers, auditors, or regulators.

Launch fast if you like. Just check first. The frameworks exist so you don't have to invent the questions from scratch.

I explain frameworks; I don't give legal counsel.