4 Practical Steps for CPAs to Start with AI

I am just a practical, old time, semi-retired non-tech savvy CPA who struggles to even use a TV remote control. Don’t waste my time with theory. Just give me answers to a few basic questions:

• What specific practical tasks can artificial intelligence (AI) do today to make a CPA’s life easier?
• What specific practical steps can a CPA do today to prepare AI for the 2025 tax season?
• How much will this save me?

CPAs understand that there is incredible potential with AI, but they don’t know what to do with it. Last year, with my frustration and passive aggressive nature in hand, I marched into the office of Chris Stephenson, our Director of Intelligent Automation, and aired my AI grievances. I realized his answers to the above questions could help a lot of CPAs.

To him, the answer to all three of the questions above lies in the following: It’s called AI Discovery. Last year he created a think tank with about 200 CPA firm professionals who overwhelmingly said, “I don’t know where to start.” Here are some practical first steps he recommends:

1. Set up an ideation session. This part is a workshop with partners and staff that gives everyone an opportunity to present their problems and ideas for possible AI solutions. Everyone on your team probably has AI ideas.
2. Initiate scoring: So, you got a bunch of ideas but now what? This is where scoring is critical. Discovery often hinges on scoring each idea based on objective measures to determine their priority. Every firm is different, but you want to score on several dimensions like which can get solved the fastest, which will have the biggest impact, which can scale for every department, etc.
3. Do technology infrastructure assessment: Ok, you know which AI projects are going to have an impact on your business but is your firm’s system even ready for AI? You’ll need to have your tech infrastructure evaluated to determine its AI readiness, and to find out if there are any shortcomings preventing you from getting started.
4. Create a roadmap: Finally, you need a clear roadmap to lay out how you’re going to deploy each solution. Do you need to buy a solution? Build an AI? How about training employees? How do you measure if your solution is having an impact?

He instantly identified five problems where AI could help:

• Collecting payments from clients
• Requesting and processing documents from clients
• Moving trial balances to workpapers
• Responding to IRS letters
• Setting up a Chatbox to handle internal queries, like HR questions

With this information, AI solutions can be created to address each of these problems. Do your own AI discovery to identify your firm’s biggest pain points and develop a plan to attack with AI.

With AI in hand, I think about CPAs increasing their bottom-line profits by looking at the following:

• The time spent on client billing analysis such as excel spreadsheets that track time, project by project, to justify hours and bills, and put that into a client letter
• Those routine IRS Notice letters written to explain why a client’s 1099-DIV did not match the tax return Schedule B
• Other mundane tasks which I used to do in public practice that bored me

Would life have been better as a CPA 45 years ago when I started in public practice if AI existed back then? Should I even consider a return to public accounting from my semi-retirement with AI in the picture?

The answer to the first question is an obvious “Yes” and to the second question an even more obvious “No.” But for those still in the game, let’s face it. It’s time to jump on the AI bandwagon to be more profitable, have less stress, have more free time and have more fun.