US audit firms and investor groups have urged the Public Company Accounting Oversight Board (PCAOB) to move quickly on guidance for the use of artificial intelligence in audits, saying existing standards need clearer interpretation as companies and auditors increasingly deploy AI across financial reporting and audit processes.
The recommendations came in responses to the PCAOB’s first request for public comment on its standard-setting and research agenda. While the respondents broadly agreed that AI should be a priority, most favoured principles-based staff guidance and research over prescriptive standards, citing the rapid evolution of the technology.
PwC said AI use by companies and auditors “should be the board’s foremost priority”. In the short term, it recommended that the PCAOB issue guidance to “raise awareness and provide clarity” on applying professional standards as AI and generative AI use expands across audits.
The firm also recommended that the PCAOB establish an AI task force, deepen engagement with firms of different sizes and coordinate with inspection personnel as it develops longer-term approaches.
Grant Thornton similarly said AI should initially be addressed through “comprehensive, principles-based staff guidance rather than through formal standard setting”. It argued that prescriptive requirements could quickly become outdated as AI technologies evolve, while suggesting future guidance to be principles-based and technology-neutral.
The firm also said the PCAOB should distinguish between AI used by financial statement preparers and AI used by auditors, since the two applications create different risks, judgments and responsibilities.
“These perspectives give rise to different risks, judgments, and responsibilities, and warrant separate consideration, particularly with respect to generative AI or AI-enabled workflows,” they said.
Baker Tilly said the PCAOB should place “greater value” on timely interpretive guidance than on new auditing standards for emerging technologies. It supported guidance clarifying how existing standards apply to audits involving emerging technologies, including companies’ use of GenAI in financial reporting processes and audit firms’ use of GenAI-enabled tools.
Focus on audit evidence, risk and human oversight
The submissions highlighted the need to clarify how existing audit standards apply when AI is used to generate evidence, perform audit procedures or assess risks.
CBIZ CPAs said the PCAOB should begin with a non-authoritative, principles-based framework and illustrative examples.”This approach would allow the PCAOB to provide timely clarity while avoiding prescriptive requirements that may become outdated as tools evolve,” they added.
The firm said the guidance should address documentation, supervision and review, prompt validation, professional skepticism, fraud, risk assessment and audit evidence. It added that the greater the effect of an AI output on audit evidence or audit conclusions, “the greater the need for documentation of tool governance, inputs, assumptions, limitations, testing, review, and corroboration.”
The Members of the Investor Advisory Group (MIAG), meanwhile, said AI and large language models were probably “too nascent” to justify PCAOB standard setting at this stage. It nevertheless backed PCAOB research and said staff guidance should remain the preferred tool as the regulator learns more about the technology.
The MIAG added that AI does not change an auditor’s responsibility to exercise professional skepticism, obtain sufficient appropriate audit evidence and apply professional judgment. “The Board’s research should therefore include examining how technology may improve fraud risk assessment and how the use of D&T may itself create or obscure risks of material misstatement,” they added.
Investors seek independent PCAOB AI expertise
The International Corporate Governance Network (ICGN), an investor-led organisation, went further, urging the PCAOB to build independent technical expertise rather than relying on the audit firms it oversees.
“The PCAOB will need to conduct its own independent research and have appropriate internal technical expertise, or access to independent third party experts,” ICGN said. It added that the regulator “must not be in a position to rely on the audit firms’ experts” and should make its own assessment of the protections investors need.
ICGN said AI and technology should be considered alongside fraud because advances in AI, without safeguards, could mask risks and lead to material misstatements. “Advancements in the use of AI and technology, without safeguards, can mask risks, leading to material misstatements in a company’s financial statements,” they added.
KPMG proposed that the PCAOB go a step further by establishing a dedicated AI task force to coordinate research, monitor developments, engage stakeholders and advise the board on future guidance and standard-setting priorities.
Audit independence rules also in focus
The comment letters also called for a review of the PCAOB’s ethics and auditor independence framework.
EY said modernising auditor independence requirements should be a high priority, with the PCAOB working with the SEC to eliminate outdated interim standards and redundant or unnecessarily prescriptive requirements while retaining investor protections.
ICGN said independence remained a key investor concern and flagged the relationship between non-audit fees and audit fees, as well as auditor tenure.
“Investors become concerned when non-audit fees climb disproportionately compared to audit fees and the length of time that the same audit firm has provided the external auditing function,” ICGN said.
PwC also said the PCAOB’s interim ethics and independence standards should be revisited in coordination with the SEC, pointing to the overlap among PCAOB, SEC and AICPA requirements and the resulting complexity.
Investors push for audit quality disclosures
The PCAOB’s investor advisory groups also pressed for greater transparency around audit quality.
ICGN ranked Critical Audit Matters (CAMs), Firm and Engagement Metrics (FEMs), Data and Technology/Fraud, Auditor Independence, and noncompliance with laws and regulations among its priorities.
The MIAG said CAMs and FEMs should be the highest priorities because they are most likely to provide investors with useful information for investment and voting decisions and have “the greatest potential to encourage audit quality competition among auditing teams and firms.”
However, the audit industry was more cautious. Baker Tilly said it did not support new requirements for firm and engagement performance metrics, arguing that the costs of collecting, maintaining, validating and reporting such data could outweigh the benefits and that metrics could be misunderstood without context.
CBIZ similarly said CAMs and performance metrics should undergo further post-implementation analysis and stakeholder outreach before any new standard-setting activity.
The submissions collectively point to a preference for a more flexible regulatory approach as the PCAOB considers its next phase of standard setting: use research and staff guidance to address fast-changing technology, reserve formal standard setting for demonstrated gaps, and ensure investor protection remains at the centre of the process.
The debate over AI, in particular, reflects a balancing act between giving auditors enough clarity to use rapidly evolving tools responsibly and avoiding rules that could become obsolete before they are implemented.
