Artificial intelligence is no longer something accounting firms are evaluating. It's already here.
Walk into almost any firm today and ask, "Do you use AI?" and you'll probably hear a variety of answers.
"Not really."
"We've experimented with ChatGPT."
"We have Copilot."
"We're waiting to see where all this goes."
Then ask a different question.
"Does your team use Microsoft 365? QuickBooks? Zoom, Teams, or Google Workspace?"
Almost every hand goes up. Over the past year, I've had this conversation with practice leaders across the country. Almost every one of them says the same thing: "We know we need something. We're just not sure where to start."
Here's the reality.
AI has already entered your firm.
Not because leadership approved it. Because software vendors have embedded it into the tools your team already uses every day.
That's the conversation I believe our profession is still somewhat missing.
What I'm hearing is backed up by the research. AI adoption across accounting firms teams has grown rapidly (more than >75% of staff use AI), yet only a small percentage of firms (<21%) have established formal governance around how those tools should be used.
That Gap Matters
Every day, firms are making thousands of AI decisions. Most of them aren't being made by partners.
They're being made by individual employees deciding whether to draft an email, analyze a spreadsheet, summarize a meeting and auto-send the notes, or connect a new application. None of those decisions are inherently wrong. But when every person creates their own rules, the firm doesn't have governance. It has inconsistency.
Without clear expectations, every employee is left making individual decisions about what information can be shared, which tools are appropriate, when human review is required, and where professional judgment begins.
That's not a technology problem.
That's a leadership problem.
The Policy We Wrote Isn't the Policy We Need
Early AI policies were built around one primary concern.
- Don't upload confidential client information into public AI tools.
- Review AI-generated content before using it.
- Use approved software.
Those were the right conversations a few years ago.
But AI has changed.
Today it's drafting emails, summarizing meetings, reviewing documents, assisting with research, suggesting journal entries, and helping analyze financial information. The next generation of AI isn't waiting for prompts. AI agents are beginning to complete multi-step tasks, interact with multiple systems, and execute work with increasing levels of autonomy.
The technology has evolved.
Many firm policies haven't.
This Is About More Than Governance
When people hear "AI policy," they often think about another compliance document.
I don't.
I think about client trust first.
Then professional responsibility.
Then liability when those things break down.
Every AI-generated response that influences client work becomes part of your firm's responsibility. Whether AI drafted the email, summarized the meeting, or prepared the first version of a report, your firm's name is still on the final deliverable.
Professional responsibility doesn't transfer to software.
It stays with us.
Clients won't care whether an answer came from AI or a human. They hired our firm, and they expect our judgment.
That's why firms need to answer questions that go far beyond whether AI is allowed.
What information should never be entered into an AI tool?
Which applications have been approved?
When is human review required?
Who owns the final work product?
What decisions should AI never make independently?
Those aren't IT questions.
They're firm leadership questions.
Start With the Foundation You Already Have
For tax and accounting firms, AI governance shouldn't be a standalone document.
We already have a framework for protecting client information. It's called our Written Information Security Plan, or WISP. The IRS expects tax professionals to maintain one. Your AI handbook shouldn't compete with it. It should live underneath it.
When AI governance is built on top of your existing security framework, it becomes part of how your firm operates instead of another standalone policy.
More importantly, your team understands that AI isn't separate from your firm's responsibilities. It's simply another way work gets done.
Build a Handbook, Not a Policy
Policies sit on shelves.
Handbooks get used.
As AI continues to evolve, firms will evaluate new tools, new integrations, and new autonomous capabilities. Your governance framework should evolve just as quickly.
It should become a resource your team references when they're trying a new feature, evaluating a new application, or deciding whether something is appropriate before they click "Send."
Because there will always be another update.
Another model.
Another AI assistant.
Another AI agent.
So Where Do You Start?
Start by defining what AI means inside your firm.
Build your AI handbook under your WISP.
Establish approved tools.
Define what AI should never do independently.
Decide where human review is mandatory.
Review it regularly, because AI won't stop changing.
Every firm already has an AI strategy.
Some firms designed it intentionally.
Others simply haven't documented it yet.
AI isn't replacing accountants.
It's changing how accounting firms operate.
The firms that thrive won't be the ones using the most AI.
They'll be the firms that build the most trust around how AI is used.
Become a member and register for this webinar today.
About the Author
Sheri Radler, CPA, CMA, CEPA leads R Accounting Group, a Woodard Top 50 CAS Practice and a 2026 Top 100 ProAdvisor Firm. She specializes in Client Accounting Services, controllership, and helping growing businesses build financial readiness through better systems, stronger processes, and practical technology adoption.
References
- Association of International Certified Professional Accountants (AICPA & CIMA). AI in Accounting Research and Professional Insights. Research highlights the rapid adoption of artificial intelligence across accounting and finance professionals.
- Wolters Kluwer. Future Ready Accountant Survey. Recent survey findings on AI adoption, technology priorities, and digital transformation within accounting and tax firms.
- Grant Thornton. 2026 AI Impact Survey. Research examining AI governance maturity, incident response readiness, and organizational preparedness for AI risk management.
- Internal Revenue Service. Publication 5708, Creating a Written Information Security Plan for Data Security and the Federal Trade Commission Safeguards Rule. These resources outline expectations for safeguarding taxpayer information and maintaining a Written Information Security Plan (WISP), which provides the foundation upon which an AI governance framework should be built.
Insightful Accountant Newsletter
Like what you're reading?
Subscribe to our FREE newsletter and we'll deliver content like this directly to your inbox.