That said, many accounting leaders are in the pursuit of preservation; keeping young professionals engaged and enticed by accounting, and promising a stable future.

The question is: how?

The profession is asking the question out loud

The concern is way past being theoretical. The Journal of Accountancy made it their March 2026 cover story, asking directly: how will accountants learn new skills when AI does the work? 

The piece reflects what's already happening inside firms. 98% of accounting professionals are using AI in some capacity, with 55% using it several times a day.

And the shift is happening at every level simultaneously, not just at the bottom of the org chart.

The World Economic Forum, meanwhile, predicts a 15% increase in demand for accountants skilled in technology and data analysis by 2027. The profession is changing shape right in front of your eyes. But that means the people coming into it need a different kind of introduction than the people who came before them.

The uncomfortable truth is that most firms haven't figured out what that formation looks like yet. Karbon's 2026 AI in accounting data shows fewer than half of firms invest in AI training for their staff, and only 21% have a documented AI policy or strategy. Adoption has outpaced intentionality. Firms are using the tools, but they haven't designed the learning environment those tools require.

What's actually at stake?

At the end of the day, what’s lost when AI handles the entry-level work traditionally slated for juniors?

1. Error exposure

Catching your own mistake (or having a manager catch it and explain why) is one of the most effective learning mechanisms in any skilled profession. When AI catches errors before a junior accountant ever encounters them, that loop closes. The mistake never happens; the lesson is never learned.

2. Pattern recognition

Experienced accountants talk about developing a feel for the numbers. It’s an instinct that something's off before they can really articulate why. That intuition is built through repetition. Hours of reconciliations. Hundreds of client files. If AI handles the repetitions, the pattern recognition doesn't develop at the same rate or in the same way.

3. Judgment under uncertainty

Much of accounting involves situations where the right answer isn't obvious, where professional standards leave room for interpretation, and where you have to make a call. Developing that judgment requires practice in lower-stakes environments before you're making those calls for clients. Without the entry-level work, there's less runway to build it.

None of this means firms should deliberately slow down AI adoption to preserve training opportunities. But it does mean firms need to think intentionally about what replaces those opportunities.

Redesigning junior roles around oversight, not execution

Rather than having junior staff do the work AI now handles, forward-thinking firm leaders are building junior roles around reviewing AI output, querying anomalies, and escalating edge cases. 

This is a fundamentally different skill set. It’s less about doing the work, and more about evaluating it. But it does develop judgment if it's structured well and not just rubber-stamping.

This is how PwC is reimagining talent development. Their plans involve junior accountants performing manager-level responsibilities within just three years. As AI automates routine audit tasks, new hires will begin their careers as reviewers and supervisors. PwC AI assurance leader Jenn Kosar explains that a “back to basics” approach to training will place greater emphasis on audit fundamentals, critical thinking, negotiation, and professional skepticism—capabilities that once surfaced only after years of experience.

Theresa Richardson, CPA, Partner and Chief Talent Officer at US firm Withum, refers to this as an ‘AI analytical mindset.’ “We must train our teams not only to use AI,” she explains, “but to critically assess and identify when AI-driven outputs may be incorrect.”

David A. Wood, the Glenn D. Ardis Professor of accounting at Brigham Young University and member of the Profession Ready Initiative Advisory Group for research, came up with an idea that takes the concept a step further.

“Students don’t remember a lot from sitting and listening to a lecture,” he said. “But if they teach somebody else, they remember a ton. So, we thought, what if we had them mentor an AI?”

What does this actually look like? Instead of reading and testing, students train a deliberately ‘ignorant’ AI bot until it can pass an exam.

“They read a textbook chapter and then have to train an AI until the AI can pass the quiz,” David explained.

And it’s not limited to technical skills. This approach is being used for skills like interviewing and other people skills.

All of a sudden, junior accountants are mirroring how they’ll work with AI in practice.

The fork in the road

There's a version of this story that's simply optimistic: AI removes the drudgery, junior accountants get to do more interesting work sooner, including the important client facing and relationship-building work, and the profession benefits. That version may well be right.

But it requires firms to actively design the conditions for it. Professional judgment doesn't develop by accident. It develops through structured challenge, reflection, and feedback.

The firms that figure out how to engineer that environment—without relying on the scaffolding they've now automated away—will have a meaningful advantage when it comes to developing the next generation of partners.

The ones that don't will find themselves with a cohort of junior staff who are very good at reviewing AI output and very uncertain about what to do when the AI is wrong.