AI Adoption in Accounting Jumped 32 Points in One Year — Where Is Your Firm?

AI Adoption in Accounting Jumped 32 Points in One Year — Where Is Your Firm?

In January 2026, CPA Trendlines published a finding that should stop every accounting firm owner in their tracks: AI adoption in the profession jumped from 9% in 2024 to 41% in 2025. That is not a prediction. It already happened. Wolters Kluwer's 2025 Future Ready Accountant report corroborates it — 72% of accounting firms now use AI at least weekly, and 35% use it daily.

Previous technology shifts in accounting — cloud software, electronic filing, QuickBooks — took five to eight years to go from early adopter to mainstream. AI ran that same distance in roughly one year. If you're a firm owner who hasn't made a deliberate AI decision yet, the question is no longer "should we?" The question is: "What's the competitive cost of where I'm standing right now?"

This piece answers that. We'll show you where your firm sits on the adoption curve, what the competitive pressure actually looks like, and — if you're still figuring out where to start — a 30-day on-ramp built for a real accounting firm, not a tech startup.

What "41%" Actually Means

The 41% figure is real, but it covers a wide range of realities. "AI adoption" in a CPA Trendlines survey might mean the managing partner has opened ChatGPT twice, or it might mean the firm runs AI-assisted tax research, automated document summarization, and AI-drafted client follow-ups every week. The number is useful as a directional benchmark — not as evidence that nearly half the profession has fundamentally changed how it operates.

Here's a more useful way to think about where your firm stands:

Non-starter. You or your team haven't used AI for work, or it's been one-off and unstructured. No workflow has changed. This is roughly 59% of the profession — but that number is shrinking fast.

Experimenter. You've tried ChatGPT or Copilot for a few tasks. It helped, sometimes. But there's no deliberate workflow, no measurement, and no accountability. Most firms who call themselves "exploring AI" are here.

Workflow user. AI is embedded in at least one repeatable process — meeting notes, draft client emails, tax research prompts, document review. Staff use it consistently. You could describe the time savings. This is the 41%.

AI-native. AI is in multiple workflows, you're measuring outcomes, and your capacity-per-staff-member looks different than it did two years ago. A small and growing minority.

The competitive gap that matters is the one between "experimenter" and "workflow user." Getting from one to the other is not a major technology project. It's a 30-day decision.

The Competitive Clock Is Real — Here's Why

When accounting firms started moving to cloud software in the 2010s, the first adopters got a few years of operational advantage before the rest caught up. The laggards eventually made the transition without losing clients over it, because clients didn't notice the difference directly.

AI is different. The efficiency gains are visible in pricing and capacity in ways that cloud software wasn't. Three specific pressures are building right now:

Price compression from faster delivery. Firms using AI complete tax returns, financial statement reviews, and audit preparation faster. When clients compare quotes and timelines, non-AI firms either match prices and compress their margins, or lose the bid. The 59% of corporate clients who told ComplexDiscovery in 2026 that outside firms "using AI" have shown them no fee savings yet — that number reflects tolerance that's still present but eroding.

Capacity advantage. An AI-native firm can handle 20–30% more client volume with the same staff headcount. That's not theoretical. CPA Practice Advisor documented it in April 2026: Steve Tonkin & Co. reduced administrative time by 20% within 30 days of deploying Digits for workflow automation. Same team, more clients. That's compounding over time.

Revenue growth correlation. Wolters Kluwer's data is direct on this: firms with a clear AI strategy are 3–4x more likely to see revenue growth than those without one. The gap between firms with an AI strategy and firms without one is measurable now, not just theoretically plausible.

None of this means your firm will lose all its clients next year if you're still experimenting. But the window where "we're thinking about it" is a defensible position is narrowing. The profession is approaching the moment where AI adoption becomes a client expectation, not a differentiator.

The 30-Day On-Ramp for a Firm Currently at "Experimenting"

If you've tried AI but don't have a consistent workflow yet, the path forward is not a technology project — it's four decisions. Here's what to do in the next 30 days:

1. Audit your existing software for AI features you're already paying for.

Before adding any new tool, check what you already have. Microsoft 365 Business Premium includes Copilot. QuickBooks Advanced has AI-assisted categorization and anomaly detection. Karbon has AI workflow features built in. Intuit's Accountant Suite has AI tax research capabilities. Many firms are already paying for AI functionality and not using it. Start there — the cost is zero because you're already subscribed.

2. Pick one high-friction, high-volume task.

The right task is one where your staff spends at least 30–60 minutes per week on something repetitive. Meeting notes transcription and client follow-up email drafting are the most reliable starting points — immediate time savings, low error risk, and easy to compare before and after. Tax question research and initial response drafts are a slightly higher-return option if your team is comfortable iterating on AI output.

Don't try to fix three workflows at once. Pick one.

3. Run a 30-day pilot with one staff member.

Assign one person to use the tool consistently for 30 days on that task. Measure what you're actually changing: track the time spent on that task before the pilot starts, then compare it after 30 days. A 15–20% reduction in time on that specific task is a realistic outcome. Document what worked and what didn't. That person becomes your in-house knowledge source for the firm-wide rollout.

4. Name the tool and set a 90-day outcome target.

Vague AI adoption is unaccountable AI adoption. By the end of the 30-day pilot, you should be able to say: "We're using [tool] for [workflow]. By [date], our time per [deliverable] is down by [X] hours per week." That specificity is what separates firms making progress from firms who say they're "exploring AI" indefinitely.

If You're Still at "Non-Starter" — The Honest Assessment

If your team hasn't genuinely used AI for work yet, there's no shame in that — but there is urgency. Here's the minimum viable starting point. It requires two hours of setup time and one week of deliberate use:

The tool: Microsoft 365 Copilot (included in Business Premium at $22/user/month, or as a $30/user/month add-on to eligible plans). If you're already on Business Premium, you're paying for it. If you use QuickBooks Advanced, the AI features are already in your dashboard.

The workflow: Client email drafting. When a staff member needs to draft a client communication — status update, question response, engagement follow-up — use Copilot in Outlook to generate a first draft. The staff member reviews and edits. This workflow doesn't require any integration, any training, or any data migration. It just requires deliberate use.

The time commitment: One hour to enable Copilot for one user and walk through the basic prompting. One hour in week one for the staff member to run through a few examples. After that, it's five minutes per email.

The success signal: Your staff member saved one hour in the first week. Not transformed the firm — saved one hour. That's the signal that tells you to expand.

The goal here is not AI at scale. It's breaking the non-starter pattern and building internal evidence that this works in your firm's actual context.

Frequently Asked Questions

What percentage of accounting firms are using AI in 2026?

According to CPA Trendlines' January 2026 analysis and the Wolters Kluwer 2025 Future Ready Accountant report, approximately 41% of accounting firms were actively using AI in 2025, up from 9% in 2024. 72% use AI at least weekly. The profession is not at the beginning of adoption — it is approaching majority adoption faster than any previous technology shift.

Is it too late to be an early adopter of AI in accounting?

The early adopter window (2023–2025) has largely closed. The current phase (2026) is mainstream majority adoption, where most peers are already using AI for specific workflows. Competitive differentiation from early adoption is less available now — but competitive risk from non-adoption is real. Firms not using AI for routine workflows are beginning to face margin pressure relative to firms that are.

What AI tools are most accounting firms using in 2026?

The most widely adopted are AI features embedded in existing platforms: Microsoft 365 Copilot, Karbon AI (practice management), Intuit Accountant Suite AI, and QuickBooks Advanced. Purpose-built tools — Basis, Digits, Black Ore, ezylia — are gaining ground for specific workflows. The lowest-friction starting point is AI meeting notes using tools like Otter.ai or Fathom, because the ROI is immediate and requires no integration work. For measurement and financial analysis, Digits and Basis are most cited in 2026 case studies.

How do accounting firms measure ROI from AI tools?

The most reliable early-stage metric is hours per client file. Track the time your staff spends on routine tasks — data entry, email drafting, meeting notes, research lookups — for 30 days before and after AI adoption. A 15–20% reduction is achievable in the first 90 days for firms using AI consistently. Beyond efficiency, watch for capacity expansion: are you handling more clients with the same headcount? That's the outcome that drives revenue growth. For budget benchmarking, see the Atlanta Fed's data on AI spending in professional services firms.

What should a 10-person CPA firm budget for AI tools in 2026?

Based on Atlanta Fed May 2026 data, the median AI spend for professional services firms is $200–500 per employee annually. For a 10-person firm, $5,000–10,000 per year covers most workflow AI needs: a Microsoft 365 Business Premium subscription (which includes Copilot), one purpose-built accounting AI tool, and implementation time. Start with tools already in your existing software stack before adding new subscriptions. If you're seeing no measurable ROI from AI tools yet, workflow integration — not more tools — is usually the fix.

---

Your Next Step This Week

Here's the one thing to do: open your Microsoft 365 admin panel and check whether Copilot is included in your current subscription. If it is, enable it for one staff member and assign them the client email drafting workflow for 30 days. If it's not included, pull up QuickBooks Advanced and open the AI features tab.

You don't need to make a budget decision, run a committee, or build a rollout plan. You need one staff member using one AI tool for one workflow, starting this week. That's how the firms in the "workflow user" category got there — not with a strategy document, but with a 30-day decision.

The profession crossed the 41% threshold in 2025. By the time it crosses 70%, clients will expect it. You have some runway left. Use it.

---

The Crossing Report covers AI adoption in professional services firms every week — written for firm owners who want the signal without the noise. Subscribe free or unlock all 10+ insights with a premium membership.