Will AI Replace Accountants? What the Data Actually Says

Type this question into Google and you get two kinds of answers. One side tells you the profession is finished. The other tells you not to worry, usually because they're selling you something. Neither side usually shows you a number.

This is the number version. Not a forecast about what AI might do in ten years, but the labor-market data that already exists: how many accountants have actually left the profession, what the government's own job-growth projections say, how fast the pipeline of new accountants is shrinking, and what CFOs report AI is doing to their headcount plans right now, in 2026, not someday.

The short version is in the next section. The data behind it is everything after.

---

The short answer

No, and the reason has less to do with AI than most of the coverage on this topic suggests.

The US Bureau of Labor Statistics projects employment of accountants and auditors to grow 5% from 2024 to 2034, faster than the 3% average for all occupations, with about 124,200 openings projected every year. That is a government labor-market projection, made with full knowledge that AI adoption in finance is already underway, and it points up, not down.

The more useful question isn't whether AI replaces accountants. It's why a role growing faster than average, with over 100,000 annual openings, feels this precarious to the people doing it. The answer is a labor shortage that started years before generative AI existed, and understanding that shortage is what actually explains what's about to change.

---

The shortage AI walked into

Start with the exodus, because it predates the current wave of AI entirely. More than 300,000 US accountants and auditors left the profession between 2019 and 2022, a roughly 17% drop from the 2019 peak, based on Bureau of Labor Statistics data. That happened before most finance teams had touched a large language model. It was driven by retirements, burnout, and pay competition from other fields, not software.

The pipeline behind it has been shrinking too, though the most recent data suggests the free-fall is bottoming out rather than continuing to accelerate. The AICPA's 2025 Trends Report puts new accounting bachelor's and master's completions at 55,152 for the 2023-24 academic year, down 6.6% year over year, a smaller decline than the prior two years (-9.6%, then -7.4%). New CPA exam candidates dropped from 42,626 in 2023 to 28,082 in 2024, then rebounded to 16,448 in just the first half of 2025. Accounting program enrollment hit 266,506 in spring 2025, up 12.4% year over year and the highest figure since 2020. Read honestly: the pipeline damage is real, and there are early signs it's stabilizing rather than continuing to collapse.

Layer the retirement wave on top of both of those numbers. Roughly 75% of licensed CPAs are within 15 years of retirement, per AICPA figures, and more than a quarter of all accountants and auditors in the workforce are already over 55.

Put the three together and the shape of the actual problem gets clear. Demand for the work is growing. The supply of people to do it has been shrinking for half a decade, for reasons that have nothing to do with AI, and a large share of the people still doing it are approaching retirement at the same time. That is the labor market AI is entering, not the labor market AI is creating.

---

What AI is actually doing to finance headcount right now

Here is where the anxiety and the data diverge most sharply.

The Consero 2026 CFO Report, surveying 102 finance leaders at PE- and VC-backed companies, found that 97% of finance functions are now using AI in some form, up from 76% just a year earlier. 76% report measurable ROI within 12 months, and only 3% remain skeptical AI will pay off at all. None of that is surprising for a technology this widely adopted this fast.

What's more relevant to the actual question in this post's title: 87% of those same finance leaders report they are hiring more finance staff even as AI scales inside their function. Not fewer. More. In a labor market that already can't supply enough qualified accountants, adding a tool that removes hours from the preparation side of the work does not translate into removing the people, because the people were never the constraint that AI addresses. The shortage was, and still is, the constraint on the other side.

That is the honest version of "we're not replacing accountants." It isn't a promise. It's what the hiring data currently shows.

---

The part that's genuinely changing, stated plainly

The honest answer isn't that nothing changes. A 2026 CFO Dive survey conducted with Oliver Wyman, covering roughly 500 global CFOs, found 64% expect a shift away from junior-level roles specifically, with 41% expecting to add more midlevel positions and 23% expecting to add more senior ones. Survey authors have described the shape of the resulting org chart as a pyramid flattening into a diamond: fewer purely entry-level seats, more roles built around review and judgment.

That tracks with what a separate, older but still relevant data point shows about where the hours currently go. A 2023 Dext survey found close to 60% of accountants report spending too much time on manual, repetitive tasks, the exact category of work that automates first. Put the two together and the pattern is coherent rather than contradictory: the roles most exposed are the ones built entirely around manual preparation, which is also the layer of work AI reaches most completely today. The roles built around reviewing that work, catching what doesn't tie, and explaining the result to someone who has to sign it are not shrinking. In most of the data above, they're the ones growing.

This is the preparer-to-reviewer shift, and it's worth naming precisely instead of softening it. A role that consisted entirely of assembling data into a workbook is a role built entirely on the part of the work that's now automatable. A role built on reviewing that output, catching the exception, and being the person who explains the number to a controller or a board is not the same role, and the data above says it's the one with a growing number of seats attached to it.

---

Why "replace" is the wrong word for what's actually happening

Put the demand side and the supply side next to each other and "replace" stops making sense as a framing.

A tool can only replace a person if there's a surplus of people to begin with. There isn't one. There's a shortage that's been getting worse since before this technology existed, a growing amount of work that needs doing, and a labor market where the honest problem for most finance leaders isn't "how do we reduce headcount," it's "how do we get this work done with the people we can actually hire." That's a capacity problem, not a replacement story, and it's the reason the CFOs in the data above are adding staff rather than cutting it.

None of this erases the real, narrower risk sitting inside it. A role that is nothing but manual data assembly, with no review, judgment, or client-facing component attached to it, is exposed, and the data on junior-role contraction says so directly. That risk is worth naming rather than smoothing over. It just isn't the same claim as "AI is replacing accountants," and collapsing the two makes people afraid of the wrong thing.

---

What this means if you run a finance team

Audit where your hours actually go before you decide anything. The manual-task number above (nearly 60%) is a survey average, not your team's number. Time your own close for one cycle and see what share of it is genuinely preparation versus review, judgment, and communication.

Hire for review, not for typing. If the roles growing in the data above are review and judgment roles, that's what your next hire should be evaluated against, not how fast they can build a workbook from scratch.

Treat the shortage as the real constraint, because the data says it is. You are not choosing between people and software in a market with plenty of both. You're choosing how to get the work done in a market that already can't supply enough qualified accountants, with a retirement wave still ahead of you.

Say the honest thing to your own team. The people most anxious about this question are usually the ones doing the most preparation work today. The data above supports telling them plainly what's actually changing: less manual assembly, more review, and a role that looks more like the one senior people already have.

<div class="cta-block">

Want to see what shifts from preparation to review on your own close? Book a pilot and we'll scope it against a workflow already on your calendar.

Want to see the mechanism first? Start free and try it against a job you can check yourself.

</div>

---

Related reading

• Will AI Replace Accountants? What Working Inside a Top-30 Firm Actually Showed — the firm-side version of this question, based on what running AI agents on live engagements inside one accounting firm actually showed, rather than on public labor-market data.
• How to Become an AI CFO: A 12-Month Roadmap
• The AI CFO Tech Stack: What Finance Leaders Are Actually Running in 2026

---

Every statistic in this piece is drawn from public labor-market data and third-party surveys, sourced individually below. None of it is Adopt's own pilot or customer data. Figures move as new reports are published; verify against the original source before citing a number from this piece elsewhere.