The tax, accounting & audit profession in the United States has lost more than one-third of its licensed workforce since 2019, with 340,000 fewer accountants working in the US now compared to 2019. Fortunately, the talent pipeline challenges in the tax & accounting profession might be abating. Recent data shows from the American Institute of CPAs (AICPA) indicates that total undergraduate accounting enrollment is up by 25% since 2023.

At the same time, however, AI continues to add a layer of complexity in two ways. One, job listings requiring AI skills for accounting positions jumped by 67%, the largest year-over-year rise; and two, a blurred definition of entry-level work is forcing the profession to confront how junior professionals will learn critical judgment skills.

Simulation replaces learning judgement by osmosis

Carl Mayes, Vice President of Ethics and Firm Quality at AICPA, runs the organization’s Professional Ready initiative, and he explains the challenge of learning professional judgment as something that even “the largest firms are trying to crack right now."

The development of judgment within the accounting profession, he explains, has historically been built on an apprenticeship model that has junior staff absorbing knowledge by osmosis while sitting alongside more experienced colleagues. Now, however, that model is being replaced by one in which hybrid and virtual work arrangements — driven by AI, automation, and offshoring of rote tasks — are the dominant methods of training. And this is forcing a reckoning with the elements of the apprenticeship model that involve implicit learning.

Mayes, along with the AICPA, believe AI-enabled simulation is an essential ingredient to the solution. Junior staff need to learn to perform the underlying work themselves before they ever supervise AI's version of it, Mayes says, adding "how do you know what good looks like if you've never done the work yourself?”

Indeed, simulations are a critical way for future accounting professionals to gain professional judgement, he explains. “If somebody can perform a large number of tax returns in a simulated environment before they ever look at a return being populated by an automation tool, they'll have a better understanding of what they're looking for." In fact, he knows of one firm that has already built a gamified internal tool, which runs staff through return after return before they touch live, automated output, Mayes notes.

New research initiative highlights broader skills gaps

At the same time, AICPA knows that it needs to be a great contributor to this solution. To help address this challenge, AICPA's Profession Ready research — under the leadership of Mayes — conducted conversations with more than 1,000 members to identify what capabilities needed to be further developed. Researchers found three consistent gaps in professionals’ knowledge and competency, including:

• Technical skills — Tax firms report that more than half of incoming staff need remediation on foundational technical topics like debits, credits, and accrual accounting.
• Professional capabilities — Critical thinking and communication as well as adaptability and a skeptical mindset were among the top skill deficits in this category.
• Technological competency — This is critical and includes having the professional judgement to know when AI is right, wrong, and what needs to be challenged.

Mayes explains that the Professional Ready initiative revealed a competency shortfall in professionals, demonstrating they had not developed "strong enough foundational knowledge in order to be able to challenge the bot.”

Recommended actions by firms to develop judgement

While academic institutions and professional organizations are part of the long-term solutions, Mayes insists that tax, accounting & audit firms cannot wait to address this issue. He notes that the most innovative firms are already taking action, by:

• Restructuring onboarding entirely — One large firm cut technical training time in half and replaced it with dedicated critical thinking exercises that were focused on analysis and the exercise of professional judgment, Mayes says.
• Using slow periods to put in exercises for professional judgment — Junior professionals need to use their down time to put in “reps via simulation," Mayes explains, adding that firms should have them looking at 1040s and 990s. They need to "go through as many audits as they can so that when they hit the ground, they can be effective."
• Focusing on professional skills — Mayes acknowledges a sea change is underway in accounting curricula because many schools are teaching concepts more broadly and going less deep into technical minutiae. In fact, one accounting program is teaching professional skills through workshops. Still, employers must continue to fill these gaps in professional skills by teaching critical thinking, communication, and ethics now rather than waiting for a redesigned pipeline.

One of the most consequential findings underscoring the urgency to think differently about how to develop judgment in the age of AI is from the Thomson Reuters' Future of Professionals Report 2026. Tax professionals surveyed say they expect the time to develop trusted judgment to compress by one year. Further, almost half of professionals (48%) say they fear that AI is having a negative impact on independent judgment development.

This insight is critical because of the professional obligation that tax, accounting & audit firms have to their clients and the overall market. “We're gatekeepers for the capital markets, and it is an incredibly important role," Mayes says. "We can't let our stakeholders down, and this is why the exercise of that professional judgment and skepticism is at the core of our function."

You can learn more about the challenges of AI adoption in the tax, accounting & audit industry here