The UK Solicitors Regulation Authority will treat AI failures as governance failures. It will assess whether firms can prove they controlled the systems delivering legal services. Supervision will focus on upstream decisions like procurement, risk assessment and data flows. Firms must show documented evidence of structured AI governance. AI issues will surface through complaints, thematic reviews, insurer intelligence, whistleblowing or routine checks. The SRA will ask why a system was selected and how risks were managed at procurement. Post-incident fixes will not replace pre-deployment governance. Firms will be judged on contemporaneous records of AI-related decisions. AI use is now expected to meet the same governance standards as outsourcing or data processing. Supervisory expectations will emerge through practice, not new rules.