A computer science graduate I know was told, by pretty much everyone, that there would be no jobs left by the time he finished his degree. AI was going to write the code. Today he has never been busier. Neither have most of the people he graduated with.

Jensen Huang put the pattern into one sentence: "The purpose of your job, and the tasks and tools that you use to do your job, are related, not the same." Tasks get taken away all the time. The work that remains gets more ambitious. Huang has watched this for more than 3 decades at Nvidia, and he expects AI agents to make people busier than ever.

I think audit is next, and I think that is good news. The purpose of an audit is an opinion someone can rely on. The tasks are tie-outs, matching, vouching and the 8-to-10-week grind of assembling evidence. The tasks are what is being automated. The purpose stays exactly where it is.

[Figure 1: The capacity paradox]

Two things that shouldn't both be true

Look at 2026 and you can watch the profession working this out in 2 directions at once.

In March, KPMG UK cut around 440 audit roles, roughly 6% of its audit workforce, most of them assistant managers. In June, Deloitte UK offered voluntary redundancy to about 175 managers and assistant managers. Both firms cited unusually low attrition.

In the same season, Advancetrack's 2026 Accounting Talent Index surveyed 500 firm leaders across the UK, US, Canada and Australia. 3 in 4 said they are turning away new work because they don't have the people to deliver it. About the same share are at or near full capacity.

The Netherlands looks the same. CBS data cited by ING shows nearly 2 in 3 accounting, legal and notary firms short of staff, and ING expects fees to rise by at least 4.5% for the 4th year running. Exact's 2024 survey found 59% of Dutch accounting firms short-staffed, and 40% said the quality of their work had dropped as a result.

[Figure 2: Market data of the capacity paradox; Data source: Advancedtrack Accounting Talent Index 2026]

Too many people and too few, in the same profession, in the same year. If capacity is a headcount question, that is a contradiction. If capacity is a question of what each auditor-hour produces, it makes complete sense. The largest firms need each head to carry more. The mid-tier cannot add heads at all. Different symptoms of the same limit.

The pipeline won't close the gap in time

The FRC's 2025 Key Facts and Trends shows UK and Irish student numbers down 0.3%, audit qualification awards down from about 1,900 to 1,800, and the number of registered audit firms down by a quarter since 2020. In the Netherlands, more people are leaving the profession than entering it, more than 1 in 4 NBA members are between 51 and 60, and student intake is expected to flatten from 2026.

These are decade-long trends. No hiring push turns them around before your next busy season, and I don't think any firm is really planning as if it could.

So where is the capacity going?

If the answer isn't more people, it has to be in how the hours already on the team get spent. In my experience, most of them go to the part of the job that needs the least training. That is where I'll pick this up in the next piece.