For more than twelve years, I’ve recruited across the Tax, Finance and Accounting market, partnering with both commercial organisations and public practice firms across two hemispheres (Europe and APAC).
During that time, the profession has navigated some of the most disruptive events in modern economic history; Brexit, the COVID-19 pandemic, the war in Ukraine, rising geopolitical instability, and growing protectionism across global markets. Each event brought hiring freezes, restructures, and prolonged talent shortages that reshaped how businesses approached workforce planning.
Now the industry faces another transformation: artificial intelligence (AI).
Unlike previous disruptions, AI has been embraced enthusiastically across accounting and finance. From a commercial perspective, the appeal is obvious. Businesses can reduce labour costs, improve turnaround times, scale more efficiently, and operate with leaner teams, all while managing increasingly complex regulatory obligations.
And in many ways, the technology genuinely delivers.
AI can already automate reconciliations, prepare workpapers, analyse transactions, draft journal entries, assist with tax research, and process vast volumes of financial data in seconds. At the same time, firms continue expanding offshore delivery models for process-driven work.
On the surface, accounting appears to be becoming more efficient than ever, but underneath those efficiency gains, a more structural problem may be developing - the profession is gradually eliminating the very roles that historically produced future experts.
That matters because while AI will undoubtedly replace parts of accounting, it is far less likely to replace deep technical judgement. In fact, the more automated the profession becomes, the more valuable genuine expertise may become.
The issue is that the pipeline producing that expertise is shrinking.
The Profession Is Removing Its Own Training Ground
For decades, accounting followed a relatively predictable development model.
Graduates entered the profession performing foundational work: tax compliance, reconciliations, audit testing, financial reporting support, and workpaper preparation. Much of it was repetitive… But that repetition served a purpose.
Those early years gave professionals exposure to transaction flows, technical standards, review processes, risk identification, materiality, and stakeholder management. Over time, repetition evolved into judgement.
The graduate preparing tax returns eventually became the tax manager advising on transaction structuring. The junior auditor became the technical partner reviewing complex accounting treatments. The assistant accountant became the CFO making strategic decisions.
The profession’s expertise pipeline relied on large volumes of junior-level exposure. Today, that layer is steadily disappearing.
AI and Offshoring Are Targeting the Same Work
What makes the current shift particularly significant is that both AI and offshoring are replacing the same category of tasks: process-driven, repeatable, rules-based work.
Historically, this was exactly the type of work performed by graduates and junior accountants. Now, firms are automating more of it, moving operational work offshore, and reducing entry-level hiring in the process.
The early signs are already visible. A 2025 HR Review article reported that all Big Four firms in the UK significantly reduced graduate intake, with KPMG reportedly cutting graduate and school-leaver recruitment by 29%. This approach has been mirrored across other international markets, including Australia.
The short-term economics make sense. Firms become leaner, margins improve, and productivity rises.
The longer-term consequences, however, are less visible.
If significantly fewer people enter the profession through traditional junior pathways, then significantly fewer people will accumulate the experience required to become high-level technical specialists later in their careers.
The impact may not fully emerge for another five to ten years, when firms begin searching for experienced experts and discover the pipeline beneath them is dramatically smaller than previous generations.
Accounting Is Still a Judgement Profession
There is a growing assumption that AI will eventually handle most technical accounting work.
But accounting is not purely a knowledge profession, it is fundamentally a judgement profession.
The hardest parts of the job are rarely the calculations themselves. They involve interpretation, context, risk assessment, commercial nuance, and knowing when something “doesn’t look right.”
You do not become an exceptional tax advisor simply by reading legislation. You become one after years of reviewing compliance work, handling edge cases, responding to regulator queries, and seeing how technical rules operate in real commercial environments.
Likewise, strong financial reporting specialists are developed through repeated exposure to real transactions, reporting issues, audit scrutiny, and stakeholder pressure. That kind of judgement is difficult to shortcut.
Tax Teams Are Already Showing the Problem
Tax is perhaps the clearest example of where this trend may lead.
Even before generative AI became mainstream, tax teams across both professional services and industry were already facing severe talent shortages. Many firms now struggle to hire professionals at the Senior Consultant to Senior Manager level who possess both strong technical capability and stakeholder management skills.
The reason is simple: high-level tax expertise takes years to build.
Historically, professionals developed those skills through heavy exposure to compliance work; reviewing returns, interpreting legislation, handling ATO interactions, responding to review notes, conducting technical research, and working through increasingly complex client scenarios.
Ironically, compliance is now one of the areas most exposed to automation, workflow software, and offshore delivery models.
The profession therefore faces a growing contradiction: the market still desperately needs highly skilled technical experts, while simultaneously reducing the number of junior professionals developing the foundations required to become them.
The Expertise Shortage May Become Severe
Accounting has always relied on a pyramid structure.
A broad base of junior talent gradually narrows into a smaller pool of experienced specialists and leaders. But if the bottom of the pyramid contracts significantly, the top inevitably shrinks as well.
That could lead to a future defined by severe shortages of technical subject matter experts, weaker succession pipelines, rising salary inflation for experienced professionals, and reduced mentorship capacity across the industry.
Ironically, AI may ultimately increase the value of genuine expertise rather than eliminate it, heightening demand for human critical-thinking skills.
As automation becomes more widespread, the professionals capable of validating, interpreting, and challenging AI-generated outputs may become disproportionately valuable.
In the future, the most sought-after accountants may not simply be the people who know how to use AI tools, they may be the people who understand accounting deeply enough to know when the AI is wrong.
The Industry May Be Optimising for the Short Term
None of this suggests firms should avoid AI or offshoring. Both are now permanent parts of the profession and, when implemented properly, can create enormous operational benefits.
But there is a difference between improving efficiency and weakening long-term capability.
At present, much of the industry appears heavily focused on cost reduction and scalability. Far less attention is being paid to whether the expertise pipeline itself remains sustainable, and that the candidate population possesses the technological skills to manage and audit their new AI colleagues.
Future experts must come from somewhere - historically, they came from large graduate cohorts spending years learning underneath experienced reviewers and mentors. If those pathways continue shrinking, the long-term pool of technical specialists may become dramatically smaller than the profession expects.
And by the time the shortage becomes obvious, rebuilding that pipeline may take another decade.