AI governance is a product decision, not just a policy one

Treating AI governance purely as a compliance checklist produces systems people route around. Treating it as a product decision — deciding upfront what the system will never do, and making that boundary visible and explainable — produces systems people actually trust.