How CPAs are putting AI to work in audits

Two years ago, CPA firms were exploring ways to use artificial intelligence in the audit. Today, they are increasingly using AI tools in everyday audit work.

To see how that shift is playing out, the Journal of Accountancy spoke with leaders at four firms ranging in size from 11 employees to more than 1,800 employees. The discussions detailed how the firms are using AI in ways that streamline audit workflows, assess risk, and review audit reports, saving time and money.

Despite the growing role of AI in audit engagements, leaders at all four firms stressed that human judgment and professional skepticism remain central to the value CPAs provide (see “AI Raises the Stakes for Auditor Judgment” on page 14).

J. Sean Spitzer, CPA, president and COO at Atlanta-based Smith + Howard, a 750-employee firm, said that robo-advisers have been used in financial planning for years to guide investing decisions, but people still hire financial advisers because they want someone who can answer their questions.

“I don’t think AI is replacing accountants,” he said. “It’s going to make us more helpful.”

HOW AI IS BEING USED

The four firms interviewed for this article identified five distinct AI use cases:

• Risk assessment and planning;
• Financial data extraction, formatting, and verification;
• Sample selection and agentic automation;
• Working paper review; and
• Report drafting.

Let’s explore each of these use cases.

Risk assessment and planning

Wells CPA, an 11-person firm that primarily serves not-for-profits, uses AI to review client financial statements, board minutes, policy manuals, and other documents to identify possible risks and plan the audit.

“AI can go through and do all of those [initial document review] processes … in a matter of two or three hours,” said Donald Wells, CPA, the firm’s president and owner. “If we had a midlevel staff person doing it, we might take a week.”

The firm performs more than 80 audits each year, so the potential time savings are significant. Wells notes that his auditors use a closed AI system, so client documents are stored securely in Wells’s IT system.

Wells noted that this use of AI requires staff to check the accuracy of the AI’s work and determine if it reached the right conclusions in identifying risks that need to be assessed.

That can prove a challenge for some staff, especially younger staff, who may be less knowledgeable or less confident in their knowledge. “[They] tend to say, ‘Hey, AI said this is a risk, therefore it’s a risk,'” Wells said. “I’m like, ‘No, we have to think it through.'”

At Smith + Howard, generative AI tools have cut research time during the planning process significantly, Spitzer said. For example, when doing research on GAAP rules, AI can quickly identify the relevant FASB standards that accountants can then apply to the audit they’re working on.

“You can just ask some of these models, ‘Where do I find the information about this specific problem,'” he said. “Instead of spending two hours flipping through guidance, you’re analyzing data. We find it allows us to do more for our clients.”

Financial data extraction, formatting, and verification

Ronald Martinez, CPA, audit partner at 1,800-employee PKF O’Connor Davies, described a custom-built agent designed to address a recurring challenge: general ledgers exported from client accounting software are often in a format unsuitable for the auditor’s use and require lengthy manual preparation before audit procedures can begin.

“Automating that preparation helps our associates focus more of their time and attention on planning analytics and risk assessment,” Martinez said.

Jennifer Wood, CPA, a partner at The Bonadio Group, said AI can handle the footing, crossfooting, and tie-out of client financial statements, freeing up staff from necessary but rote tasks. Wood is the assurance service line leader at the firm and was involved in developing the AICPA’s Dynamic Audit Solution framework for using artificial intelligence (AI) in audits.

“That could take them two to three days to do manually, where now we can use these AI tools that are designed to do stuff like this,” Wood said. As a result, she said, the staff accountant’s role on that task shifts “from a doer to a reviewer.”

Sample selection and agentic automation

Some of the most tangible benefits firms are seeing from AI involve analyzing large datasets and identifying exceptions that might otherwise go unnoticed.

At The Bonadio Group, an engagement team executing a data query with AI identified a handful of employees who had authorization to post journal entries but had done so fewer than three times during the year. The finding prompted the client to tighten controls and reduce software-license costs for employees who rarely used the application.

In addition, Wood and Martinez described building or deploying AI agents that go a step further than analysis, such as selecting audit samples from client transaction populations.

Wood pointed to Microsoft Copilot’s agentic tools as one example: “Copilot has come out with some really cool agentic AI tools where you can kind of program and have the tools do things like selecting your sample for you,” she said.

She added, however, that some of The Bonadio Group’s biggest wins have come from independently built, in-house agentic tools rather than off-the-shelf platforms.

At PKF O’Connor Davies, Martinez said the firm has built several custom Generative Pre-Trained Transformers (GPTs) and agents, like an agent that formats general ledgers, to “reduce … routine work so our professionals can apply their experience where it has the greatest impact.” The tools are vetted by the firm’s Digital Transformation Committee before they are rolled out firmwide.

“Internal AI champions have been important,” Martinez said. “They help colleagues build comfort and confidence as the tools are introduced.”

Working paper review

Separately from anything that accesses client data, Wells described using AI as a “first reviewer” of the audit team’s own working papers — checking documentation for completeness and clarity before a human reviewer ever sees it.

The AI review is mandatory. No client-facing document is reviewed by a partner or manager until the auditor responds to all the AI’s comments. Wells said the process has shifted his own review time away from catching errors and toward judgment calls specific to the client.

Report drafting and disclosure checking

Wells CPA also uses AI to check draft audit reports for internal consistency — flagging inconsistent naming conventions or places where a report doesn’t “tie to itself.”

Wells’s team also uses a vendor-supplied checklist tool that can review financial statements and determine whether they’ve met all the disclosure requirements. “It will tell you, ‘AI found this disclosure, AI did not find this disclosure, are you OK with not having this disclosure?'” he said.

Some of Wells’s nonprofit clients have board members who are attorneys accustomed to reviewing 10-Ks and 10-Qs. They sometimes send back audit reports, or IRS Forms 990, Return of Organization Exempt From Income Tax, that Wells’s team has prepared, with extensive markups. With AI functioning as another reviewer, there’s less red ink now.

SECURITY IS STILL CRITICAL

Wells, Wood, Spitzer, and Martinez noted that ensuring that client data and confidential information remain subject to tight controls is important.

Wells CPA, for example, runs on a closed Thomson Reuters platform, CoCounsel. Wells notes that despite his enthusiasm for AI’s potential, he bars his professional staff from using ChatGPT because of concerns regarding confidential client data.

At Smith + Howard, the firm has chosen an enterprise version of Claude so it can be comfortable using it on confidential client data.

“We have an approved list of tools that can be used across much of our work,” Spitzer said. “Claude enterprise has enterprise-level security, which we got comfortable with. Copilot has Microsoft security behind it.”

JUNIORS LEARN JUDGMENT

AI adoption is also driving a broader shift in what junior staff accountants do.

Wells’s firm is training junior staff not to defer blindly to AI output, reinforcing that the conclusion and the work product are still their responsibility. Wells predicted that some entry-level fieldwork auditing tasks will be fully automated within a few years, pushing junior staff into review roles earlier in their careers.

Wood noted that the potential of AI to reduce and automate rote tasks is changing how auditors work and even how they make decisions. “The days of SALY, same as last year — those are gone,” she said.

Smith + Howard’s Spitzer said that despite narratives about AI destroying jobs, his firm is hiring more people even as it adopts AI tools.

“When you make someone four times more efficient, you can do more. I want more of those people,” Spitzer said. “We’re hiring. We’re not cutting because of AI. We expect to be able to do more for our clients in the future. It’s an exciting time to be in our industry.”

---

About the author

Mark Tosczak, MBA, is a freelance writer based in North Carolina. To comment on this article or to suggest an idea for another article, contact Jeff Drew at Jeff.Drew@aicpa-cima.com.

---

LEARNING RESOURCES

AICPA ENGAGE 2027

Advanced auditing and artificial intelligence will be on the agenda at the biggest event in the accounting profession. For more information or to register, scan the QR code or go to aicpaengage.com.

June 7—10, ARIA Resort & Casino, Las Vegas

CONFERENCE

Digital CPA 2026

Join CPA.com in San Diego and online from Dec. 6—9 to discover emerging technologies, unlock actionable strategies, and collaborate with peers who are designing a new era of accounting.

CONFERENCE

Josi — Gen AI tool built for auditing and accounting professionals

Josi is a frontier model LLM with access to AICPA and PCAOB standards, FASB/GASB codifications, and industry guides (including disclosure checklists). Josi keeps your data and your clients’ data secure, answers the hard questions, and does the heavy lifting on document review, research, and accelerating your team. Use the tool, but apply your professional judgment.

AI TOOL

Certified Information Technology Professional

The AICPA’s Certified Information Technology Professional (CITP) credential represents the intersection of technology and financial reporting. CITPs understand how information technology, accounting, and finance interrelate. Choose from three pathways to earn the credential.

CREDENTIAL

For more information or to make a purchase, go to aicpa-cima.com/cpe-learning or call 888-777-7077.

---

MEMBER RESOURCES

Community

Rise2040

A global community of finance and accounting professionals working together to co-create a bold, unified vision for the profession’s future.

Articles

“5 Human Competencies CPAs Need in the AI Age,” JofA, May 1, 2026

“Audit Report Card: More Internal Audit Teams Suffered Cuts in 2025,” JofA, March 24, 2026

“How Will Accountants Learn New Skills When AI Does the Work,” JofA, March 1, 2026

“COSO Creates Audit-Ready Guidance for Governing Generative AI,” JofA, Feb. 26, 2026

“How AI Is Transforming the Audit — and What It Means for CPAs,” JofA, Feb. 1, 2026

Podcast episode

“The Auditing Standards Board’s Priorities for 2026 and Beyond,” JofA, May 7, 2026