Insights — Research Paper
Abstract
Most AI governance frameworks in financial services are existing model-risk frameworks with AI-specific language layered on top. This substitution misses the real governance question: who is accountable when an AI-assisted decision is challenged, and can that accountability be evidenced.
The paper recommends mapping, decision by decision, what independent judgment a human reviewer must exercise before an AI-assisted recommendation becomes final — a decision-rights exercise, not a technology one.
Read the full paper for the complete argument, evidence, and recommendations.
Ravi Sharma — Principal, Regulatory Strategy