AI can be helpful in accounting, but firm leaders must approach it skeptically and cautiously. There seems to be an increasing amount of over-promise and under-deliver going on in the market. You risk sinking large sums into technology that delivers only frustration.
I saw a commercial online recently for a company unveiling a consumer-ready, fully autonomous AI robot designed to live in your home. The ad showed it carrying groceries and serving coffee like something out of a sci-fi movie. When I returned to my computer and tried to get our firm’s new AI software to pull some simple information from some meeting minutes for summary … fail. It missed a clear reference to a “legal settlement,” even though I specifically asked it to find evidence of litigation. So, what’s going on?
The reality is AI can be helpful in accounting, but firm leaders must approach it skeptically and cautiously. There seems to be an increasing amount of over-promise and under-deliver going on in the market. Without the right approach, you risk sinking large sums into technology that promises “an ROI of thousands” but delivers only frustration.
As a leader in a regional accounting firm of about 130 employees, I’ve seen how AI is reshaping how we work. It’s not just about flashy tools or dashboards, it’s about using technology thoughtfully to make our teams more efficient, our audits more insightful, and our client experience stronger. But before diving headfirst into every new system that lands in our inbox, we must pause and ask, what do we actually need?
Many regional firms (ours included) tend to want to move too quickly. The market is bursting with AI tools, each claiming to “revolutionize” your workflow. It’s easy to get swept up by a polished demo and think, we need this! But too often, firms realize months later that the real issue wasn’t a lack of software, it was a lack of clarity about what problem they were trying to solve.
Before exploring AI, first take a hard look at your firm’s true operational needs. Where are the bottlenecks? What manual tasks drain staff time? Which areas create frustration for clients or risk for the firm? Once those pain points are clear, you can better evaluate whether an AI tool truly aligns. Without that clarity, you risk what I call “distractive innovation” (coined from real-world experience), which is chasing every new tech trend instead of making meaningful progress.
Audit systems are a great example. About a year ago, our firm had to make a change in audit systems due to the discontinuance of a vendor relationship. During that change, it became clear that the trends are moving to smarter, more data-driven audits. We had to make a choice: do we stay more traditional and let those systems develop, or move forward at the front of that wave? Fast-forward 12 months, and now we are in the middle of an audit software change from Caseware to Inflo, which promises the ability to use big data to create audits that are both more efficient and more effective. The transition has been what we expected it to be—difficult—but once our teams adapt, we’ll incorporate full sets of client data and let the system go to work. Most importantly, we made this change because we had a need, and that clarity helped us stay focused through the transition.
Automated document processing is another area drawing attention, and this type of software seems to be everywhere. Machine learning models can now read and classify PDF documents, extracting key details like vendor names, amounts, and dates with a “reasonable” (said skeptically) level of accuracy. However, these systems can be costly, and our experience hasn’t matched the vendor’s ROI promise. We aren’t backing out yet, but we are pushing the vendor hard to prove their value! If you’re exploring this space, pilot it thoroughly using your own documents and stress test the system before committing.
Of course, no discussion of AI in accounting is complete without mentioning ChatGPT and Microsoft Copilot. These tools are becoming staples across firms, helping to streamline document analysis and uncover useful insights. But open-ended systems like these carry risks, especially around overreliance and data confidentiality. Firms should set clear expectations with all staff: skepticism and verification must guide every use. Blind reliance on AI output isn’t acceptable, and staff should document how they verify what the system produces. However, given their accessibility and low cost, these tools can be a logical first step for firms beginning their AI journey.
So, will AI replace accountants? Not anytime soon. Firms that use it effectively will undoubtedly outperform those that don’t. The key, though, is balance—embracing innovation while staying grounded in your firm’s strategy. The best results come when technology serves a purpose, not when it’s adopted out of fear of missing out. Start with clarity, pilot with diligence, invest in your people, and ensure each system truly supports how your firm delivers value.
Ian F. McDowell, CPA, is principal, audit and assurance, with S.R. Snodgrass PC. He assists public companies in their SEC regulatory filings, navigating business combinations, and establishing public and private stock offerings. He is a member of the American Institutes of Certified Public Accountants and Pennsylvania Institute of Certified Public Accountants (PICPA), the PICPA Financial Institutions Committee, and a PICPA Board member.
S.R. Snodgrass will have a presenter at PICPA’s Insurance and Financial Institutions Conference on Nov. 12 and 13, 2025. Check out this year’s agenda and sign up today.
Sign up for PICPA's weekly professional and technical updates by completing this form.
Statements of fact and opinion are the author's responsibility alone and do not imply an opinion on the part of the PICPA's officers or members. The information contained herein does not constitute accounting, legal, or professional advice. For actionable advice, you must engage or consult with a qualified professional.
