The accounting AI software market was worth roughly $6.68 billion in 2025 and is projected to reach about $37.6 billion by 2030, a compound annual growth rate near 41%. That growth rate would be remarkable in any sector. In a profession often described as conservative, late-adopting, and resistant to change, it’s the most interesting number in the room.
The numbers around the people doing the work tell a more measured story. The AICPA & CPA.com 2025 AI in Accounting Report found 85% of accountants feel excited or intrigued by AI, with the top stated benefits being speed and efficiency (also 85%), error reduction (68%), and task automation (65%). But only 37% of firms invest in AI training, and the US Bureau of Labor Statistics still projects 5% job growth for accountants and auditors from 2024 to 2034, faster than average, with around 124,200 openings every year. The market for the software is sprinting. The market for the people is walking. The gap between those two paces is where the next two years play out.
Where adoption sits, with the firm-level evidence
The clearest measure of how fast firms are moving comes from the Wolters Kluwer Future Ready Accountant 2025 report, which tracked AI adoption inside firms going from 9% to 41% year over year. Most of that movement is in client-facing firms, particularly mid-sized practices using AI for document review, ledger reconciliation, and first-pass analytical work.
The AICPA report shows the same direction: accountants list speed and efficiency as the dominant benefit, then error reduction, then automation of repetitive tasks. The work being targeted first is, predictably, the work humans were never the best fit for: matching transactions to source documents, summarising long contracts, drafting standard correspondence, parsing tax notices. None of that is the part of the job that builds a career.
What’s notable is that the 37% training-investment figure is not keeping up with the 41% adoption figure. Most accountants now using AI are learning it on the job, from prompts shared in Slack channels and LinkedIn posts. That works for individual experimentation. It does not produce the audit-trail discipline a profession built on documentation actually needs.
What the BLS projection actually says
The BLS Occupational Outlook Handbook projection of 5% growth for 2024 to 2034 is one of the more useful sanity checks against headline doom narratives. There were roughly 1.6 million accountant and auditor jobs in the US in 2024. Five per cent growth on that base, plus replacement demand, produces about 124,200 openings each year for the next decade.
The BLS rationale matters as much as the number. They cite continued business activity, more complex tax and regulatory environments, and the need for human judgement on close calls in audit and advisory work. Software has been automating accounting tasks since the 1980s. The job kept growing because the volume of business activity grew faster, and because regulatory complexity created more, not less, work for trained humans. AI is the latest layer of automation, not the first.
The skill premium is real, but it’s not what marketing claims
Most “save 7 to 9 hours a week” figures in the accounting AI marketing space are not from controlled studies. The closest peer-reviewed evidence on AI productivity for adjacent knowledge work is the Harvard Business School and BCG study on consultants given GPT-4, showing roughly a 25% reduction in completion time within scope, and the GitHub Copilot RCT showing 55.8% speedup on a benchmark task. Both are different work to accounting, but they bracket the realistic gain. Most accountants who adopt AI for routine drafting and summarisation will see something in the 20 to 40% range on those specific tasks. That is meaningful. It is not the entire week.
The career premium shows up clearly in job postings. Roles asking for AI literacy in accounting and finance have multiplied in the last two years, and the salary differential between an accountant who can run a tax research workflow through Claude or ChatGPT and one who cannot is widening at the senior end.
What the next 12 months look like
Three movements are visible in the data. First, the firm-adoption rate will keep climbing. Going from 9% to 41% in a single year does not flatten quickly, especially when the underlying software market is compounding at 41%. By the end of 2026, the share of firms using AI for at least one billable workflow should be in the high fifties or low sixties.
Second, the training gap will start to close, but unevenly. The largest firms will lead, midsize firms will follow with formal programs, and the long tail of small practices will rely on individual learning. That is the same pattern every previous accounting technology has followed (think cloud-based ledgers in the early 2010s).
Third, the conversation will shift from “does AI replace accountants” to “what does the audit and review process look like when first-draft work is machine-generated.” That is a regulatory and professional-standards question, not a productivity one, and the institutions that govern the profession move slowly. Expect AICPA guidance and IRS positions on AI-assisted preparation to evolve through 2026 rather than land all at once.
The accountants who treat AI as a research and drafting partner, with their judgement and signature still attached to the output, are the ones whose remits expand. The market growth is not going to wait for anyone to be ready.