AI won’t replace auditor judgment but it will raise the bar

If AI adoption in your firm has left you feeling uncertain about where your experience fits in, you’re not alone.

The narrative in the market is that AI is transforming the profession at high speeds, yet for auditors on the ground, these tools often feel disconnected from the reality of the work – not to mention their expertise. If anything, tacked-on AI assistants can be more of a nuisance than a help.  

The real opportunity lies not in learning new technology, but developing the skills to evaluate, challenge, and apply professional judgment to AI outputs and processes. As AI takes on more mundane, repetitive work, your informed perspective [or “human analysis”] becomes more valuable, not less.

Jessie Kanter, a partner who leads assurance innovations at Citrin Cooperman, addressed the topic during a webinar titled “AI in Audit: What’s Real, What’s Hype and What Actually Matters.”  

“I think that is the biggest issue that we are facing as a profession right now,” she said. “It is the number one topic of discussion. What are the skills that our people need in the future and how do we get those skills for those people?”

The auditors who thrive won’t be the ones who simply use AI tools. They’ll be the ones who know how to make AI useful via human intervention.

Data analysis: the foundational skill

AI can process entire transaction populations, identify anomalies and surface patterns at unmatched scale. But it cannot determine whether they matter. That power rests with the auditor. Those who know what to do with AI outputs are will add real value and make an impact at their organizations.

Technology and AI literacy: understanding, not just using

An auditor who can use an AI tool will follow its outputs. An auditor who understands how it works will question them. In an AI-enabled engagement, that’s the difference between having an answer and defending a conclusion.

“I think that’s going to be a skill that people are going to need,” Kanter noted. “Understanding what the reasoning means, being able to read through it and understand, ‘Did that AI reach a conclusion that is supportable, and that I, the human, can accept?’”

This doesn’t require auditors to become data scientists. It means auditors need to learn enough about how a tool reaches its conclusions to evaluate whether those conclusions are defensible and traceable to a reliable source. In practice, it means identifying when an output lacks evidence, recognizing where human review is required, and analyzing the processes that led to the output.

Professional judgment and ethical frameworks: essential human accountability

When routine tasks are automated, judgment doesn’t lose importance. It becomes the job. Auditors must evaluate the AI’s work, apply professional skepticism, and reach conclusions they are prepared to sign off on.

As Jason Bradley, VP of AI and Methodology Innovation at Caseware, said during the AI in Audit webinar: “LLMs are fantastic at taking huge amounts of data and synthesizing and creating things. They are significantly less good at nuanced judgment.” That gap, between synthesis and judgment, is precisely where the auditor’s value lies.

It also means auditors need to understand which tools have been approved for which use cases, what level of human review each stage requires, what to do when an AI output looks wrong, and how to document their review in a way that satisfies professional and ethical standards.

“We won’t look at anything that doesn’t cite the source and give us a clear audit trail back to where the agent or the LLM got the result from,” Kanter noted. “If we don’t have that traceability, we’re not even going to look at it.”  

The right technology to apply evaluative skills to AI workflows

To develop AI-era skills, auditors need tools that can be challenged, reviewed and traced back directly to the source. Otherwise, there is no way to evaluate whether an output is defensible in an engagement.

Caseware Verity operates as an inline intelligence layer built directly into the audit workflow, not as a tacked-on assistant. It maintains engagement context from planning through reporting, drawing on firm methodology, supporting documentation and professional standards to help auditors surface risks, analyze documentation, and exercise judgment with confidence.

From executor to evaluator: the auditor’s greatest strength in the AI era

The good news is that auditors are not starting from scratch in learning how to work with AI.  

The skills that make a great auditor – skepticism, critical thinking, professional judgment, accountability – are the same skills that matter most in an AI-enabled audit and assurance practice. AI may change how the work gets done, but it does not change who is responsible for reaching a defensible conclusion.  

As repetitive work becomes increasingly automated, auditors who can evaluate evidence, challenge assumptions, and exercise sound judgment are even more valuable to their teams and clients.

Caseware Verity meets you inside the engagement, helping auditors evaluate AI-generated insights without sacrificing the rigour the profession demands. Put AI to work in a way that strengthens, rather than replaces, the professional judgment at the heart of every audit.

Download the eBook to learn how to reskill your audit and assurance team for the AI era: Reskilling Your Audit and Assurance Team for the AI Era.

‍