AI assistants are increasingly delivering answers about products, services, and organisational obligations that differ from approved internal documentation. These externally generated representations bypass existing content controls and create a misstatement layer - a parallel communication surface that may expose enterprises to regulatory, legal, reputational, safety, and compliance risks. This briefing offers: • A rigorous taxonomy of misstatement types • A likelihood × severity risk matrix with calibration guidance • An inherent vs. residual risk analysis against existing controls • A comprehensive menu of control strategies (preventive, detective, corrective, compensating) with cost/complexity ranges • A regulatory and legal context showing possible exposure across sectors • An ownership and accountability model mapped to typical enterprise functions • A sector-based severity map to prioritise resources • A decision-tree for risk triage and action sequencing • Proposed success metrics to track control effectiveness This document is structured to support Risk Committees, Compliance, Legal and AI Governance teams in assessing whether misstatement risk is material, and to guide the design of proportionate controls.