Three Rules for Junior Accountants Competing With AI

The modern aspiring accountant’s rising challenge—besides clearing the usual exams, data clerk load, and meeting deadlines—is AI. Not in the sense that it is taking their jobs, if anything, there is a staffing shortage in junior accountant roles, but in the sense that it is absorbing the workload that junior accountants need to build their skills. This comes at a time when the accounting industry already struggles to attract and retain junior professionals.

Although there has been some recovery in the talent pipeline, the incoming junior accountants might find that their usual workload of reconciliation, categorization, document collection, and basic tax prep has been taken over by AI.

This creates a conundrum for juniors on how to progress in a profession where one’s knowledge is built through years of exposure to repetitive accounting tasks. On a broader level, it damages the credibility of the accounting profession if future accountants with little real exposure to the actual math reach senior positions and exercise poorer judgment because they are used to AI taking over the reins. Early evidence from a Stanford and MIT study on the use of AI by accountants shows that less experienced accountants are more likely to rely on uncertain GenAI outputs, whereas a senior accountant is more likely to intervene. The problem is not simply that AI is taking over junior work. It is that the profession still needs juniors to become experienced accountants capable of catching this newly introduced tech.

So, how do junior accountants deal with this new competitor? I propose 3 rules. But first, it helps to understand what has changed.

The collapse of the training pyramid

71% of tax professionals are on board with the use of GenAI in the field. The areas they find it most useful in are tax research, tax preparation, tax advisory, accounting/bookkeeping, and document summarization. Because of this, seasoned accountants might not see the merits of juniors at the moment, as the data entry tasks that would once be relegated to juniors can simply be done by AI, and at a much faster rate. The merits of AI in accounting cannot be underestimated either. A study found that GenAI use reallocated 9% of accountant time from routine data entry toward high-value tasks, increased ledger granularity by 12%, and reduced monthly closing time by 7.5 days. For firms used to working late hours during peak tax season, this is an obvious improvement.

But the reality remains that accountants are leaving the profession faster than new ones are entering. In 2025, the AICPA reported that 75% of current CPAs were eligible to retire within 15 years. This is coupled with a contraction between 2018 and 2023 of graduates, but the numbers have been rebounding after 2024. Still, the current BLS estimate is that the U.S. will need roughly 115,300 accountant and auditor openings filled each year between 2025 and 2035, mostly because existing workers retire or move into other occupations. That translates to around 2.1 openings for every accounting degree awarded.

While AI has been overall received positively in the accounting field, 50% of finance organizations report insufficient human talent and skills as holding back the adoption of the technology. Accounting is losing workers and graduates at the same time that AI is changing what remaining accountants are expected to do. The changing landscape means that firms are now looking for people who understand both accounting and AI, but the training pyramid is strained due to junior work being done by AI. That leaves the much-needed juniors entering the talent pipeline in an awkward position: they are in an industry that needs them but does not have the work to train them towards building the robust senses of a senior professional.

Rebuilding the training pyramid

The argument then is not to preserve inefficient work because it is helpful as training. The evidence of AI boosting productivity for a field already under strain is clear. In fact, individual accountants appear to be adopting AI much faster than many firms formally integrate it. The question is, how does the accounting industry rebuild the training pyramid to robustly accommodate the juniors it direly needs?

A better alternative is to provide juniors more supervised responsibility, review AI outputs, and investigate exceptions. Training may rely more heavily on simulations, historical cases, and deliberate exposure to unusual scenarios instead of spending years on routine work. But mentorship is the key to this. The industry is already based on a strong culture of codified standards, education, certification, and continuing professional development. A new training manual does not need to be rebuilt from scratch.

Some firms have already started redesigning junior work. For example, PwC is training junior accountants to be like managers within three years. Some disagree with this method, but the industry is in a state of flux as firms figure out how to adapt. The more important questions remain for those entering the field on how to prepare themselves for the current state of affairs.

The 3 rules

Junior accountants cannot personally control what a firm’s training program looks like. But they can strategize how to make themselves more valuable than software by incorporating 3 rules.

Rule 1: Build client fluency early

Client relationships and professional judgment remain heavily dependent on human skills. Traditionally, junior accountants built those capabilities gradually after spending years on technical and repetitive work. As AI automates more of that entry-level work, firms are having to expose juniors to judgment, advisory work, and client interaction much earlier. The sooner junior accountants understand how clients think, how their businesses operate, and what actually matters to them, the sooner they can add value beyond producing technically correct outputs. This means sitting on calls, understanding the client’s industry and how they make money, and asking questions to understand previous judgments. Communication, critical thinking, influencing, and business partnering are vital in standing out in this era of AI-assisted workflows.

Rule 2: Learn context engineering

“Context-engineering” is not the same as “prompt-engineering.” The latter is about improving how you phrase the question to the AI; the former deals with controlling what information the AI has access to before answering. In the context of accounting, it means giving AI the right client data, policies, contracts, prior decisions, and current standards before asking it to handle tasks like tax analysis, audits, reconciliations, or technical accounting. Of paramount importance are a client’s industry, history, and type. Learning to identify and supply that information forces junior accountants to understand the business behind the numbers rather than simply accept an AI-generated answer.

Humans never leave the loop. Every deliverable is reviewed by a professional, and every tax form still carries the name of the individual who signs it. AI doesn’t do the math. Calculations are performed using deterministic rules defined by accounting standards and IRS requirements. Current discussions of AI in audit emphasize data quality, source reliability, output evaluation, and documentation for any technology-assisted analysis. The exact qualities that context engineering operationalizes.

Rule 3: Chase mentorship actively

Previously, an apprenticeship meant sitting near someone doing the work. Since that proximity is disappearing, along with the routine work itself, mentorship has to be requested more proactively. Programs like EY’s 2026 Career Residency are a great example of initiatives firms are taking to provide some of the crucial mentorship that juniors are losing access to. By combining hands-on client work, skills development, and coaching, future professioınals will have the necessary wherewithal to meet the AI challenge head-on as well as progress their accounting careers without feeling hopeless.

However, not all firms provide such opportunities. In that case, juniors must seek out opportunities. For example, one can volunteer to sit in on client conversations as a silent observer and ask the senior afterward what they were listening for and why they handled the conversation the way they did. Another example is to review past cases and trace the decision-making. Lastly, juniors can also look beyond their firms and join state CPA societies, alumni networks, industry associations, and professional communities for guidance that an AI agent simply does not have access to. With the workload of most senior professionals, it would be naive to assume that most would have the time to take someone new under their wing. This is why mentorship must be pursued from different communities rather than just one person.

Conclusion

Accounting is one of the few industries where AI has not yet actually eroded the talent pipeline. However, the deployment of GenAI to optimize workflows and fill in the current shortage could manifest in the production of a future accounting workforce that may rely too much on AI’s output rather than their own critical thinking. In order to create a workforce that can meet the challenges of AI head-on and build appropriate accounting skills, junior accountants must proactively build their communication skills, learn how to deploy AI to their advantage, and pursue mentors and residency programs that provide client exposure in the right dosage to build a skilled professional. On the supply side, accounting firms must also create the required programs to meet juniors halfway and ensure the talent pipeline does not constrict once more.

ABOUT THE AUTHOR:

Slava Orekoff is co-founder and CEO of Numica. A serial entrepreneur and global executive, he spent more than a decade at SAP, rising to senior vice president, and has more than 25 years of experience across business, accounting, and technology. He first built accounting software at age 17, holds an Executive MBA from Stanford Graduate School of Business and master’s degrees in Physics and Economics, and mentors at the Stanford Venture Studio.