The International Auditing and Assurance Standards Board (IAASB) has published new guidance explaining how existing quality management standards apply to artificial intelligence tools used in audit and assurance engagements.
Published on 11 November 2026, the 24-page document, Guide 1: Common Conceptual Foundation, is the first instalment of a four-part series.
The guidance focuses on AI tools used to design or perform procedures for obtaining audit and assurance evidence. It addresses responsibilities at both firm and engagement level, the lifecycle of AI tools, and considerations for identifying and assessing quality risks.
What risks does the IAASB identify?
The guide discusses several potential risks associated with AI-enabled audit tools:
Hallucinations: AI tools may generate fabricated information or produce outputs that appear credible but are inaccurate, incomplete or unsupported.
Automation bias: Auditors may place excessive reliance on AI-generated results, particularly when outputs appear plausible.
Limited transparency: Some AI tools may be difficult to test or explain, making their outputs harder to evaluate and challenge.
Missing information: AI tools may omit important details without indicating that information is missing. The guide gives an example involving the extraction of non-standard terms from contracts.
Changes in performance: AI models may change over time, potentially affecting the validity of earlier evaluations and approvals.
Widespread errors: Tools deployed across multiple engagements may repeat inaccurate outputs or inappropriate actions before problems are identified.
Who remains responsible?
The IAASB confirms that audit firms and engagement partners retain their existing responsibilities when using AI, even when the technology performs a substantial part of an audit procedure.
Audit firms remain responsible for determining whether tools are appropriate for their circumstances, including when those tools are supplied by a global network or external technology provider.
The guide also addresses human oversight. It explains that effective involvement may depend on whether individuals have the necessary competence, sufficient time, relevant information and authority to challenge or override AI outputs.
Further guidance planned for 2027
The IAASB plans to publish the remaining guidance in stages:
January–February 2027: Firm-level quality management before AI tools are deployed.
April–May 2027: Engagement-level use and oversight.
July–August 2027: An integrated edition covering all four guides, including monitoring and remediation after deployment.
The timetable is subject to IAASB approval.
The IAASB emphasises that the guides are non-authoritative and introduce no new requirements. They explain how firms can apply existing standards, particularly ISQM 1 and ISA 220 (Revised), to AI-enabled technological tools.
Guide 1 is currently a pre-final version and may be revised as the remaining guidance is developed.
Related reading
PwC Tells PCAOB to Make AI in Audit Its “Foremost Priority” — PwC urges the US audit regulator to address automation bias, unreliable audit evidence and the application of existing standards to AI.
Big Four AI Hallucinations: KPMG, Deloitte and EY’s Report Failures — Examines fabricated references, unsupported claims and weaknesses in reviewing AI-generated material.
PwC to Train 85,000 Staff in Critical Thinking as AI Reshapes Work — PwC’s $100 million programme to strengthen professional judgement and employees’ ability to challenge AI-generated conclusions.
Big Four Back Plan to Ease US Audit Quality-Control Rules — Examines proposed changes to the PCAOB’s QC 1000 standard and the Big Four’s positions on audit quality-management requirements.
Want to stay up to date on all things Big Four around the world?
Check out the News section and subscribe to Big4News for weekly deep dives and briefings.
About the author: Claudine Cassar is the founder and editor of Big4News, covering audit, consulting, regulation and governance across Deloitte, PwC, EY and KPMG.



