Few technologies have generated as much excitement—and as much promise—for accounting firms as artificial intelligence (“AI”). The potential to streamline audit execution, reduce hours, and enhance firm profitability is real and already being realized. However, AI does not simply change how audits are performed; it fundamentally alters how firms must think about oversight, responsibility, and quality management. As regulators sharpen their focus on AI‑enabled audits, firm leadership must move beyond adoption and address a more complex challenge: establishing clear and scalable AI governance. This article outlines why AI governance is now a strategic imperative for accounting firm leadership. As discussed in JGA’s article What Regulators Expect to See When AI is Used , inspectors do not evaluate AI tools in isolation. They evaluate whether the engagement team obtained sufficient appropriate audit evidence, exercised professional skepticism, and applied appropriate supervision and review when AI was used. Those expectations are grounded in existing auditing standards and apply regardless of whether AI was used for risk assessment, testing, or documentation support. Against that backdrop, AI governance is not simply about approving tools or managing technology risk. It is about ensuring the firm’s system of quality management supports consistent, supervised, and well-documented use of AI that aligns with audit objectives and withstands inspection scrutiny. When firms treat AI as an IT matter, governance discussions tend to center on 1) Data security, 2) System access, 3) Vendor due diligence, and 4) Infrastructure controls. Those topics matter—but they are only the baseline. Inspectors do not evaluate whether AI systems are well engineered; they evaluate whether AI enabled audit work complies with standards, supports professional judgment, and is governed within the firm’s system of quality management. In short, AI governance is a firmwide audit quality issue, not a back office technology function. Using AI does not change the auditor’s responsibilities. Requirements still apply when AI is used for 1) Audit evidence, 2) Professional skepticism, 3) Supervision and review, 4) Engagement partner accountability and 5) Firm level quality controls. From an inspection standpoint, AI introduces new audit quality risks, including: Over reliance on automated outputs Reduced professional skepticism (automation bias) Inconsistent application across engagements Insufficient documentation of judgment Lack of transparency around how conclusions were reached These are not IT risks—they are audit quality risks. AI Touches Nearly Every Component of a QC System Under modern quality management frameworks (including PCAOB QC 1000 , AICPA SQMS No. 1, IAASB ISQM 1), AI affects nearly every component of a firm’s QC system, not just technology or data governance. 

In our work with firms, we have seen a clear shift in how monitoring and remediation are viewed under modern quality management frameworks. They are no longer treated as retrospective compliance exercises. Instead, engagement deficiencies are increasingly used as meaningful inputs into an ongoing, risk-based system designed to identify issues early, address them thoughtfully, and reduce the likelihood of recurrence. Regulatory messaging reinforces this evolution. Oversight bodies are signaling a shift in focus from isolated engagement outcomes and more on whether firms have a system of quality management that consistently detects quality risks, responds appropriately, and demonstrates that remediation is working in practice. Based on our experience, while individual engagement deficiencies remain important, the more critical question is becoming how firms analyze, respond to, and learn from those issues over time. Engagement Deficiencies Are Signals, Not Endpoints Engagement deficiencies can surface through many channels, including pre-issuance reviews, internal inspections, post-issuance reviews, peer reviews, and regulatory inspections. Regardless of source, firms benefit most when these findings are evaluated through a consistent quality management lens. In practice, we encourage firms to look beyond whether a single engagement fell short . The more meaningful consideration is whether the deficiency points to potential weaknesses in governance, methodology, training, supervision, resourcing, or monitoring activities. We often observe that when issues are quickly labeled as engagement-specific, without assessing whether they reflect broader quality risks, valuable insight is lost. Modern quality management frameworks are designed to use these signals to strengthen the system, not simply close individual findings. What Effective Monitoring and Remediation Looks Like in Practice Firms that navigate this environment effectively tend to apply a disciplined and repeatable approach when deficiencies are identified. Based on our experience supporting firms across a range of practice areas, several elements consistently make a difference: Assess whether the issue may be systemic Recurring observations across engagements, service lines, or time periods often indicate system-level risk. Similar documentation gaps, inconsistent application of methodology, or supervision challenges rarely arise in isolation. Perform meaningful root cause analysis Effective root cause analysis typically moves beyond surface explanations. Firms benefit from evaluating whether policies and procedures were designed appropriately, implemented as intended, and supported by sufficient training, time, and resources. Design remediation that directly responds to the quality risk Remediation is most effective when it is clearly linked to the underlying risk. Depending on the circumstances, this may include enhancements to methodology, targeted training, revised review requirements, or changes to engagement acceptance, staffing, or oversight processes. Validate remediation through timely monitoring Implementing corrective actions is only part of the process. In our experience, firms are most successful when they also confirm that remediation operates as intended. Follow-up monitoring performed early enough to prevent recurrence is a critical component of this step. Failure to validate remediation remains one of the most common and consequential weaknesses we observe across firms. Case Study: When Remediation Is Not Validated In one situation we encountered, a firm identified engagement deficiencies through post-issuance reviews. The issues mirrored observations that had previously been noted during peer review and were communicated as having been addressed by the group responsible for report issuance. However, responsibility for validation was not clearly assigned, and no follow-up procedures were performed to evaluate whether the revised processes were effective. Subsequent post-issuance reviews, triggered by an organizational change, revealed that similar and additional deficiencies had re-emerged. From a quality management perspective, this was not an engagement execution failure. It reflected a breakdown in monitoring and remediation. The firm had information indicating quality risk but did not adjust its monitoring activities to confirm that remediation was working. Viewed through a system lens, this represents a system-level deficiency rather than an isolated engagement issue. Quality Management Applies Across All Engagement Types Modern quality management frameworks apply across a firm’s assurance and attestation practice, including private company audits, public company audits, SOC engagements, nonprofit audits, and other services. Deficiencies identified in any practice area may signal broader weaknesses in: Governance and leadership Methodology and training Monitoring activities Remediation processes In our experience, firms struggle to maintain an effective system of quality management when certain practices are treated as exempt from system-level evaluation. Key Takeaways Engagement deficiencies are inputs into the system, not endpoints. Recurring issues often indicate systemic quality risk. Remediation should be validated, not assumed. Monitoring activities should evolve as risks emerge. Quality management applies across all engagement types. Firms that treat monitoring and remediation as a continuous feedback loop, rather than a periodic exercise, are typically better positioned to improve engagement quality and respond to evolving regulatory expectations. Looking for an independent perspective on whether engagement deficiencies have been fully addressed? Based on our experience working with firms across assurance and attestation practices, Johnson Global Advisory supports clients by performing independent reviews, validating remediation efforts, and strengthening monitoring processes. If you would like support refining policies, training, workflows, or documentation standards, or would benefit from an objective assessment ahead of regulatory, peer, or internal inspections, contact your JGA audit quality advisor to discuss your needs.