Accounting firms are no longer waiting to see whether AI will matter. They are buying tools, running pilots and giving employees access to AI-enabled software.
The harder question is what happens next.
In the latest episode of the INSIDE Public Accounting Podcast, Executive Director Chelsea Summers talks with Donny Shimamoto of Intraprise TechKnowlogies, who helped develop the AI questions in IPA’s 2026 Information Technology Survey. Their conversation moves beyond whether firms are “using AI” to examine what adoption actually looks like inside a firm and what leaders should expect from the investment.
The survey captures a profession well into experimentation but still early in transformation. More than a third of participating firms describe themselves as implementing AI initiatives, while just 3.2% say AI is fully integrated into how the firm operates.
This gap shapes much of the conversation.
Individual Productivity Is Only the Beginning
For many firms, AI adoption still means helping individuals work faster. An employee may use an AI assistant to draft an email, summarize a meeting or automate part of a task.
Those uses can save time, but Shimamoto distinguishes them from changing how an entire engagement, department or service line operates.
A firm can have plenty of employees using AI without having redesigned its workflows around the technology. And because most accounting firms will depend on their existing software vendors to bring AI into those workflows, firm leaders also need to look past the label on the product.
Shimamoto’s advice is to get specific: What is the tool actually doing? Where does it fit into the work? What changes because the firm has it?
Decide What Success Looks Like Before the Pilot
The same problem appears when firms try to determine whether their investments are working.
Nearly 43% of participating firms said they had not yet seen a measurable impact from their AI tools. Others reported improvements such as faster turnaround, better employee experience and additional capacity.
For Shimamoto, that means firms may need to broaden what they consider a return.
An AI investment does not have to eliminate a position to create value. If it reduces busy-season hours, gives experienced professionals more time for advisory work, helps a team meet deadlines or reduces engagement risk, those outcomes matter too.
But firms have to decide which result they are trying to produce before the experiment begins. That gives leaders a better question than “Did AI save us money?” What was this pilot supposed to improve and did it?
Governance Comes Before the Purchase
One of the sharper tensions in the survey is the pace at which firms are buying technology compared with the work happening around it.
Nearly three-quarters of participating firms purchased software or modules for AI-enabled functionality. Investments in data governance, cybersecurity enhancements and more advanced training were far less common.
Shimamoto argues that the order should be reversed.
A firm should know how it will evaluate vendors, protect confidential information and govern the technology before it brings a new tool into production. That becomes especially important as AI moves beyond personal productivity and begins touching client work.
And the issue exists even when leadership has not formally rolled out an AI product. Staff may already be using AI for meeting notes, emails or other everyday work.
The governance conversation cannot wait for the firmwide implementation.
Not Every Firm Needs an AI Architect
The resources available to address those questions vary dramatically by firm size. A large firm may have dedicated technology, data and innovation teams. A midsized firm may have an IT leader already responsible for everything from cybersecurity to getting new laptops into employees’ hands.
Shimamoto does not argue that every firm needs to respond by hiring an AI specialist.
Instead, he ties that decision back to the firm’s strategy. If AI embedded in existing accounting applications can meet the firm’s needs, outside expertise or vendor support may be enough. If the firm intends to build its own capabilities or sees technology as a source of competitive advantage, bringing that expertise in-house may make more sense.
The staffing decision follows the strategy, not the technology trend.
The Bigger Change Has Not Arrived Yet
For all the discussion about AI replacing accounting jobs, that is not what most firms in IPA’s survey are reporting.
Almost 8 in 10 participating firms said AI has had no noticeable impact yet on staffing and hiring. Shimamoto sees that as further evidence that much of the profession remains in the early stages of adoption.
The larger opportunity may be what firms do with the capacity AI creates. Can professionals spend more time with clients? Can the firm take on work it previously lacked the capacity to serve? Can technology improve the employee experience or reduce risk?
Those questions move AI out of the technology budget and into the firm’s broader operating strategy.
Know Where You’re Going Before You Spend More
Asked what one thing he would tell every managing partner to do next, Shimamoto returned to something much less flashy than a new AI tool: assess where the firm stands now and decide where it wants to go.
Without that starting point, firms risk adding pilots, products and subscriptions without knowing what they are collectively supposed to accomplish.
Or, as Shimamoto put it: “Wandering costs money.”
For firms heading into another round of AI planning and purchases, that may be the more useful measure of progress: not how many tools the firm has adopted, but whether leadership knows what it wants those tools to change.
