SUMMARY Auditors conducting risk assessment can use Generative AI (GenAI) to analyze large volumes of complex information from multiple sources. However, the technology underlying GenAI tools often struggles to retain and learn from feedback. This commentary proposes that auditors can address these limitations by organizing audit-relevant documents in knowledge graphs whose structured context persists across GenAI interactions. Auditors using knowledge graphs gain traceable reasoning paths that support explainability, constrain GenAI retrieval to relevant evidence, and reduce computational costs by limiting the documents processed. This commentary also discusses how knowledge graphs can be integrated into existing audit workflows under current auditing standards. JEL Classifications: M42.
SUMMARY The accounting profession is undergoing a period of rapid change in education requirements for CPA licensure. Many states and jurisdictions are reducing minimum required education for CPA licensure from 150 to 120 hours (Hood 2023; AICPA & National Association of State Boards of Accountancy (NASBA) 2024, 2025). However, little research has examined the relation between education level and CPA performance in practice. In this article, we summarize the published study by Barandi, Campbell, Guo, and Notbohm (2025), which finds a positive association between audit engagement partner graduate education and audit quality. The authors suggest that graduate programs enhance critical thinking, communication, and technical skills that are vital to auditing. As alternative pathways to licensure gain traction, these findings highlight the practice implications of graduate education. Public accounting firms, audit committees, and regulators could consider these results when making decisions and assessing the long-term effects of CPA licensure policy changes. JEL Classifications: M42.
On March 31, 2026, the Public Company Accounting Oversight Board (the Board or PCAOB) issued a request for comment on its strategic plan (PCAOB 2026). This comment letter summarizes the views of the participating members of the Auditing Standards Committee of the Auditing Section of the American Accounting Association. Our comments include responses to the questions listed in the request for comment. Our key recommendations are to prioritize inspections-particularly through expanded support for smaller firms, return to the supervisory model, advance standard setting in areas such as technology, ICFR, and complex estimates, and enhance transparency across inspections, quality control reporting, and PCAOB operations.
This study examines whether large language models can predict critical audit matters (CAMs) from management discussion and analysis (MD&A) disclosures in 10-K filings. We tested four models, OpenAI's O3-Pro and O3-Mini High, Anthropic's Claude Sonnet 4, and Google's Gemini 2.5 Flash, using zero-shot prompts against auditor-identified CAMs across 141 firm-year observations in technology, finance, consumer discretionary, and healthcare. We benchmarked predictive capability matching auditor CAMs and generative capacity, identifying audit-relevant risk matters from MD&A. O3-Mini High demonstrated the strongest predictive performance, whereas Gemini 2.5 Flash underperformed in both areas. Technology achieved the highest prediction accuracy, whereas healthcare ranked last. We also examined how management tone and algorithmic bias jointly affect prediction outcomes. These findings position artificial intelligence (AI) as a valuable screening tool to support auditor planning, while reinforcing that professional judgment remains essential to final CAM determination.
In 2022, the PCAOB undertook an interim analysis of the newly issued AS 2501 and AS 1210 using comment letters and surveys/interviews with auditors, financial statement preparers, and audit committee members. We surveyed 58 valuation specialists for use in a comment letter because they are a critical component of the fair value accounting and auditing processes. They suggest that the new requirements generally led to increased (and possibly enhanced) communications between valuation teams and other constituent groups. This additional communication seemed to increase engagement hours, likely resulting in higher one-time fees. There appear to be additional costs, and the benefits and efficiencies did not necessarily offset them. It does not appear that valuation and audit teams encountered significant challenges implementing the new requirements, but reporting entities likely did. There was consensus that the new requirements contributed to improved audit quality but perhaps exacerbated the shortage of specialists.
The Public Company Accounting Oversight Board (PCAOB) recently added inventory to its standard-setting agenda (PCAOB 2023b) to address recent economic developments and concerns that inventory audit standards have not significantly changed since 1939. Existing inventory standards primarily address inventory observation and inventory quantities. We provide insights into inventory audit issues by examining Accounting and Auditing Enforcement Releases (AAERs) from 2000 to 2022. We supplement this by reviewing critical audit matters (CAMs) from 2019 to 2023 to examine inventory audit risks and problem areas identified by auditors. The most common inventory issue in CAMs and AAERs involves obsolescence. Although many AAERs involve pricing and observation issues, these are rarely identified in CAMs. Our research highlights inventory areas where frauds and other significant misstatements have occurred. Audit practitioners can use these results to communicate these significant risks and revise audit procedures to address these risks.
Prior accounting research in the healthcare field has paid little attention to how regulatory auditing innovations, such as artificial intelligence (AI) algorithms, can influence audit outcomes. Although these innovations may improve efficiency, they may have other consequences. This article summarizes the findings of a field study by Akinyele, Baudot, Koreff, and Sutton (2025), who examined the use of an AI-based auditing tool in the United States healthcare sector. The authors found that the use of this tool can generate debates about the legitimacy of health provider claims, especially when auditors rely heavily on algorithmic decisions. When paired with an incentive-based contract, these tools may be implicated in decision-making process, and even limit healthcare options. This research underscores the importance of balancing technological auditing innovations with professional judgment to ensure effective oversight and service delivery in the healthcare sector.
In this article, we provide a practitioner summary of the paper “Auditor perceptions, reactions, and responses to PCAOB inspection feedback” (Tegeler, Brown, and Downey 2025a) which finds that when auditors perceive the feedback source (inspectors), message, and their firm’s support more positively, they react with more satisfaction and motivation to use inspection feedback. These positive reactions in turn enhance auditors’ desirability to improve audit quality and inspection risk, result in more job satisfaction, and less turnover intent. In contrast, less positive reactions lead to more impression management (i.e., managing inspection risk in excess of audit quality). We summarize key findings and discuss practical implications to enhance audit practitioners’ inspection interactions and processes.
To gain insights into external factors that impact auditors' ability to meet regulatory requirements and thus audit quality, we review the existing audit literature on individual auditor mindset through three well-accepted mindset theories (Nolder and Kadous 2018; Gollwitzer 1990, 2012; Dweck 2006). Our paper reveals that the mindset of the individual auditor is a potential factor impacting their ability to meet regulatory requirements that are intended to promote audit quality. Implications of these findings for practitioners are discussed.
: This paper summarizes a three-part study by Jefferson, Andiola, and Hurley (2025) that examines two practical strategies (microbreaks and supervisory support) to reduce auditors' fatigue and improve audit quality during busy season. First, a field study provides evidence that public accountants take fewer microbreaks during busy season, when fatigue is significantly greater and they need them the most. However, the authors find that those who do engage in more frequent microbreaks or receive higher levels of supervisory support can mitigate their daily busy season fatigue, and pairing microbreaks with support reduces fatigue the most. Second, a follow-up experiment provides evidence that even just a one-minute microbreak reduces fatigue enough to noticeably improve auditors' error detection. Third, interviews with firm senior leaders provide perspectives on implementing such strategies. Collectively, the study's results suggest that practical solutions exist to combat the effects of mental fatigue on audit quality during busy season.
In this article, we provide a practitioner summary of the paper "Auditor perceptions, reactions, and responses to PCAOB inspection feedback" (Tegeler, Brown, and Downey 2025a) which finds that when auditors perceive the feedback source (inspectors), message, and their firm's support more positively, they react with more satisfaction and motivation to use inspection feedback. These positive reactions in turn enhance auditors' desirability to improve audit quality and inspection risk, result in more job satisfaction, and less turnover intent. In contrast, less positive reactions lead to more impression management (i.e., managing inspection risk in excess of audit quality). We summarize key findings and discuss practical implications to enhance audit practitioners' inspection interactions and processes.
Staff auditors frequently email clients to conduct evidence inquiry. A recent article by Bhattacharjee, Moreno, and Wright (2023) investigates whether two characteristics of email communication, client response time and message processing fluency, influence staff auditors' judgments. The current article summarizes the study for practitioners, which finds that auditor skepticism is reduced when the client provides a more fluent email response compared with a less fluent response in a moderate response time. In addition, results show auditor skepticism is reduced when the client provides a more fluent email response in a moderate response time than an immediate response time. Implications for practice as well as practitioner and audit firm takeaways are discussed.
Generative AI and edge AI are reshaping external audit by enabling fullpopulation analysis, streamlining evidence gathering, and enhancing risk assessment. This article presents a practical roadmap grounded in a five-level AI maturity model, articulated by industry leaders, to depict what is operational today versus what is still emerging. Through firm-level mini-cases, we illustrate how auditors are using AI to draft memos, automate compliance checks, and structure unstructured data. The discussion also addresses real-world challenges, such as data security, professional accountability, and evolving skill sets, and offers actionable strategies to support responsible implementation. By bridging technical innovation with practitioner needs, this paper equips audit professionals to adopt AI tools with clarity, confidence, and governance.
Over the past three decades, the practice of voluntarily obtaining independent assurance for sustainability disclosures has grown significantly. We are now on the verge of a monumental shift in sustainability assurance due to emerging mandatory assurance requirements introduced by the Securities and Exchange Commission (SEC) adopting rules to standardize climate-related disclosures, the state of California passing three climate disclosure bills (SB253, SB261, and AB1305), and the EU passing the Corporate Sustainability Reporting Directive. This paper provides an overview of the three regulations, including their current implementation status; identifies four key elements of the sustainability assurance requirements within them; and discusses the challenges and implications for practice.
The Securities and Exchange Commission (SEC) has recently called on companies to disclose more robust forward-looking information, particularly during periods of uncertainty. Such information is used not only by investors but also by auditors when evaluating complex estimates and going concern assessments. This article summarizes "How Does Management Voluntary Disclosure Behavior Influence Auditors' Judgments?" by Hillison and Vittori specialIntscript which examines how management's choice to voluntarily disclose forward-looking information, such as earnings forecasts, affects auditor skepticism of that information. In an experiment, the authors find that auditors evaluating goodwill impairment are more skeptical of forward-looking information that has been voluntarily disclosed (versus mandatorily disclosed or held privately) due to diminished trust in management. This heightened skepticism leads to increased planned testing, even when the forward-looking information aligns with prior year trends, which could discourage future voluntary disclosure. The study's findings have potential implications for both audit quality and efficiency.
In this article, we provide a practitioner summary of our paper "Who Rewards skeptical behavior, regardless of the outcome, appear to develop staff that are more likely to detect and convey fraud red flags to their superiors. In BLS, we first identify audit supervisors who are more likely to reward appropriate skepticism. We then investigate which personality traits, knowledge, and incentives are associated with supervisors who reward appropriate professional skepticism even when no misstatement is identified. We find that trait skepticism, especially suspending one's judgment, drives the evaluations of professional skepticism in our setting. Also, we observe that when supervisors believe that their own audit partner will view the skepticism favorably, they "pay it forward" by rewarding their own staff who engage in skepticism.
Professional skepticism is a fundamental yet complex aspect of auditing. To understand how auditors exercise skepticism in practice, Xu, Yang, and Fukofuka (2023) conducted qualitative research based on stories provided by auditors. The study identified that professional skepticism is like a sensemaking process. It is triggered and unfolded through auditors' actions and interactions with their audit teams and clients. These findings highlight the importance of understanding how auditors act, converse, and interact with clients to effectively exercise professional skepticism, which has important implications for auditing firms and practitioners.
This article provides a practitioner summary of the research study titled "Auditor use of benchmarks to assess fraud risk: The case for industry data" (Brazel, Jones, and Lian 2024 (BJL)). Auditors perform preliminary analytical procedures to identify risks that financial statements are materially misstated due to fraud. Via a survey of practicing auditors, BJL find that auditors rely heavily on prior year balances and relations within the client's financial data (e.g., ratios) as benchmarks when developing expectations during planning. Meanwhile, the empirical analyses of BJL reveal that, when identifying fraud risks, benchmarks derived from industry data, nonfinancial measures, and cash flows outperform both prior year balances and relations within the client's financial data. The industry benchmark was the top performer. The article provides examples to demonstrate an approach that practitioners can use to identify fraud risks via industry data.
This article summarizes "How do financial executives respond to the use of artificial intelligence in financial reporting and auditing?" (Estep, Griffith, and MacKenzie 2024; hereafter EGM). EGM survey financial executives about their perceptions of AI in financial reporting and experimentally examines how they would incorporate AIgenerated information when resolving proposed audit adjustments. EGM find that financial executives do not have negative perceptions of AI. Further, in a hypothetical scenario, they find that when a financial executive's company uses AI to help prepare an accounting estimate, financial executives book larger audit adjustments if auditors use AI to audit the estimate than if auditors do not use AI. If a financial executive's company does not use AI, auditors' use of AI has minimal influence on financial executives' decisions. This paper provides insights for practice based on EGM's findings.
In this paper, we examine the challenges in auditing information technology (IT)-enabled financial reporting by analyzing PCAOB inspection reports for the years 2019-2024. Despite increased regulatory efforts to enhance audit quality, recent PCAOB inspection findings still reveal recurring deficiencies in public company audits, suggesting diminished audit quality. Our review of PCAOB inspection reports indicates that audit deficiencies related to IT and accounting information systems, such as deficiencies in testing automated controls and using system-generated data as audit evidence, persist and remain challenging to address. These findings are critical as more companies adopt advanced technologies in their financial reporting processes. This paper provides valuable insights into how evolving IT may heighten the complexity of audit engagements and proposes recommendations to enhance audit quality in this rapidly evolving IT environment.
