An AI governance audit finds that an organization's AI risk assessments are performed once at model inception and never updated. What is the MOST significant risk this practice creates?

B is correct because AI risk is dynamic: models drift, use cases expand beyond original scope, regulations evolve, and the threat landscape changes. A one-time risk assessment becomes stale almost immediately and fails to capture these evolving risks. Periodic or event-triggered risk assessment updates are essential to maintaining a current view of AI risk. Methodology inconsistency (A) is a quality concern that is orthogonal to assessment frequency. Computational cost (C) of risk assessments is negligible and not a relevant risk. Documentation compliance (D) is a minor administrative concern compared to the substantive risk management failure.