FRC Publishes First AI-Focused Audit Quality Review, Flags Risks in AI-Assisted Audit Evidence

What happened

The Financial Reporting Council (FRC) published the findings of its first dedicated thematic review into the use of artificial intelligence tools across UK audit firms on 24 April 2026. The review, which examined AI adoption practices at the Big Four and several mid-tier firms over the past twelve months, found that while AI is delivering genuine efficiency gains in areas such as journal entry testing, document review, and risk assessment, governance and quality control frameworks have not kept pace with the speed of deployment.

The FRC's review identified several specific concerns. Chief among them was the use of large language models to summarise and interpret audit evidence without adequate human verification — a practice the regulator described as creating 'an unacceptable risk of material misstatement going undetected.' The report also flagged inconsistencies in how firms document their use of AI tools in audit working papers, making it difficult for engagement quality reviewers and the FRC's own inspectors to assess whether sufficient professional scepticism was applied.

The regulator stopped short of issuing binding new rules but set out a series of expectations for firms. These include maintaining clear audit trails showing where AI was used and what human oversight was applied, establishing firm-wide AI governance committees with partner-level accountability, and conducting regular validation of AI model outputs against manual benchmarks. The FRC indicated it will incorporate AI-specific assessment criteria into its next round of Annual Quality Reviews starting in autumn 2026.

The review builds on the FRC's December 2025 discussion paper on technology in audit and aligns with the ICAEW's recently published AI assurance framework. FRC CEO Richard Moriarty said the review was 'not about slowing innovation' but about ensuring that 'the profession's embrace of AI enhances rather than undermines audit quality and public trust.'

What this means for your business

For audit firms — from the Big Four to smaller practices — this review represents a clear signal that the FRC expects AI governance to be formalised and auditable before the next inspection cycle. Firms that have been piloting AI tools informally or allowing individual engagement teams to experiment without central oversight should treat this as an urgent call to action. Establishing an AI governance committee, documenting AI usage in working papers, and creating validation protocols are no longer aspirational — they are baseline expectations.

The implications extend beyond audit. Accounting and advisory firms using AI for tax analysis, due diligence, or financial modelling should anticipate similar scrutiny from their respective regulators. The FRC's emphasis on traceability and human oversight mirrors the direction of travel across UK professional services regulation, including the SRA's March 2026 generative AI guidance for solicitors and the FCA's consultation on AI model risk management. Firms that build robust AI governance frameworks now will be better positioned not only for regulatory compliance but also for winning client confidence in an environment where AI-generated professional advice is under increasing public and institutional scrutiny.

For professional services leaders, the practical takeaway is threefold: audit every AI tool currently in use across your firm, ensure each has a documented approval and oversight process, and train engagement teams on the regulator's expectations around professional scepticism when AI is involved in forming conclusions. The FRC has made clear it will be looking for evidence of all three in its upcoming inspections.