Who Pays When AI Gets It Wrong: The Liability Gap in Licensed Work

As AI enters licensed professional work, the biggest unresolved risk is who bears liability when AI-informed decisions cause harm.

A commercial insurance operator raised a liability question that courts have not yet answered. When AI assists professional decisions in fields requiring licensing, such as insurance, medicine, law, accounting, and engineering, and those decisions cause injury or damages, it is unclear who is responsible. The professional, the AI provider, or both may share fault. This risk will likely be shaped by case law, but no courtroom precedent exists yet. That leaves firms exposed during a legal vacuum, with coverage and contract terms not built for the scenario. Beyond liability, the operator flagged workforce dynamics. In male-dominated fields, men and women appear to be questioned differently about their AI use, suggesting adoption friction is partly social rather than technical. The operator also pointed to a practical skills gap around prompts, specifically how they are created and used effectively, indicating that value depends on training as much as tools. For executives, the signals are concrete. Map where AI touches licensed decisions, review professional liability coverage for gaps, and add liability allocation clauses to service contracts before courts impose their own answers. On the workforce side, set consistent expectations for AI use across roles and invest in role-specific prompt training rather than generic demos. The through line is accountability. As AI moves into regulated professional work, the open question of who answers for its outcomes carries the heaviest cost and regulatory weight.