A consequence: AI without governance costs control of client risk

Legal Futures reported on October 4 that a third of law firms were already using AI tools without a formal policy governing how the tools are used, what data enters them, or who is accountable when something goes wrong. The report places governance, not experimentation, at the center of the operational question: an AI system can be added to legal work before the firm has decided who owns its failures. That is the event here, and its consequence is direct. AI without governance leaves control of client risk undefined.

That matters because human responsibility does not disappear when the work is delegated. The Solicitors Regulation Authority says AI can improve efficiency and services when used appropriately, but that recent cases show the importance of human oversight and governance arrangements. Legal Futures makes the same division more sharply: AI can do the work, but it cannot inherit the lawyer’s duty. The practical implication for an AI-native legal service is not that automation is disqualified. It is that verification, escalation and accountability have to be designed into delivery before the service can be judged by speed or cost. Bad legal work still produces consequences, and a policy gap makes it harder to identify who was supposed to prevent them.

The report does not establish that every firm using AI lacks control, nor does it show that an unmanaged tool has already caused a particular client loss. It establishes a narrower proposition: Legal Futures reports that a substantial share of law firms has no formal framework covering use, data or accountability. That is a governance condition, not proof of a failed matter. The stronger reading would require firm-level policies, an identified decision-maker, documented review and an observable consequence when those controls fail. The New Economics of Legal Review, reported by Legaltech News, describes AI as providing speed, scale, consistency and computational power while attorneys provide context, skepticism, judgment and accountability. That distinction explains why a firm’s AI deployment cannot be assessed by the system’s output alone. A workflow that produces polished work without showing who verified it has not demonstrated accountable delivery.

The next fact to watch is not another announcement that a firm uses AI. It is a published or filed operating framework that names which tasks the system may perform, what client data may enter it, who reviews the output, and who remains accountable when the review fails. A policy that merely permits AI would not settle the question. The useful signal would be a matter-level workflow that connects a human decision-maker to an auditable check, especially where confidentiality or legal judgment is involved. Legaltech News reports that AI agents disrupt the traditional workflow in which an attorney inputs information, software processes it and the attorney reviews the result, with privilege and confidentiality obligations running through the attorney’s conduct. We will treat a documented control that survives an actual matter, complaint or court filing as progress. Until then, the governance gap is not an administrative footnote; it is a cost imposed by the service itself.