AI in Accounting: What Every Forward-Looking CPA Needs to Know

Five or ten years ago, AI was the domain of Fortune 500 companies who poured billions of dollars into researching and developing the technology. Today, there’s so much that can be done with free and low-cost ai-driven tools — and the possibilities are increasing every day. The benefits of AI technology are now readily available to any size business, to help streamline processes, increase accuracy and deepen financial insights.

Yet there is also a lot of uncertainty around the role of AI in the accounting industry, with people wondering if it will decrease the number of available positions — or eliminate the need for accountants all together. I recently heard Barry C. Melancon, President of the AICPA, offer this response to those fears: AI won't replace accountants. But it will replace accountants who decide not to embrace AI.

AI is already in a lot of the things that we're doing in the accounting profession: anyone who uses cloud-based accounting software is benefiting from artificial intelligence. The real opportunity now comes from being able to utilize AI, leverage it, and be on the front end of it — versus just using AI-powered tools that come on the general market. Accountants who make strategic use of ai will give themselves a competitive advantage.

This article will discuss some of the ways accounting industry experts are putting AI to use currently, and insights into the future of AI technology in the industry:

AI can help:

• Provide client insights from financial data

• Minimize repetitive tasks

• Conduct initial research

• Streamline internal workflows

• Develop dynamic, in-house knowledge repositories.

Automation Technology: RPAs Versus APIs

There are so many different ways to leverage AI and automation. To get started, here are two technologies commonly used by accountants to boost their clients’ automation: Robotic Process Automation and Application Programming Interfaces.

1. Robotic Process Automation

Robotic process automation (RPA) eliminates many time-consuming accounting processes, including the review of journal entries, account reconciliation or testing of access control. Benefits of Robotic process automation include fewer mistakes and better compliance, in addition to cost savings.

2. Application Programming Interfaces

Application Programming Interfaces (APIs), are protocols that share data between programs, for example sending data from QuickBooks to Xero. APIs add functionality to the original program, using it to do simple things like update excel spreadsheets in real-time, or even more complex tasks (powering forecasting and visualization tools). CPA firms can use APIs to import client data into accounting software and customize ai applications for clients.

How AI Systems Can Improve Client Services

Facilitate Comprehensive, Real-Time Data Analysis

The biggest impact of AI will be the push to move accounting professionals away from bookkeeping and towards advisory.

By inputing client historical data along with new transactions (after stripping out any confidential information), AI tools can recommend the entry information including: input of client information, vendor names, and the accounts, and even the confidence level of each entry.

While the setup takes time, once it’s up and running, monthly financial statements might come together in just 15 minutes. That leaves significant time left over to dig into any anomalies and provide high-level insights into financial planning or cost-saving opportunities. All of a sudden you have a sustainable process for even really high expertise areas like fractional CFOs.

Another area of development is invoice recognition: thanks to natural language processing and machine learning algorithms, invoices arrive in an inbox and get recognized and recorded. As Kyle Geers, CEO and co-founder of Zeroed-In Consulting, explains, “They’ve already started putting rules to automatically categorize a transaction. It’s still up and coming. But as it gets refined, I can see that changing things for a lot of accounting firms, where you’re not spending 10 to 20 hours sorting transactions. You’re doing some review, but then you have more time to analyze what this means, more of that strategic side.”

“I don't necessarily think that AI is going to replace the accountants within the next year or two,” says Kyle. “But the accountants that adopt new technologies are going to succeed. The amount of time that it's going to take to complete a task is going to go down considerably for a lot of these things. Firms that stick with the status quo and say, ‘Hey, let's get someone to spend eight hours on this,’ that not a very sustainable business model.”

Just be careful: Human error will decrease as AI tools improve — but not disappear. Even if you’re able to reduce an hours-long task to a few minutes, don’t forget to budget time for a thorough review so that quality doesn’t suffer.

Fully Automate Workflows

As AI-powered tools become more sophisticated, one important distinction to keep in mind is the difference between automation and automation-assisted. Automation-assisted accounting work is fairly common: for example, scheduling a recurring journal entry.

Jesse Rubenfeld, Founder of FinOptimal, explains what full automation looks like: “If you're selling software and you get a closed deal in HubSpot, we connect our data wrangling tool that grabs data from HubSpot. We apply some logic that's set up front. We send the invoice and we calculate and book the automated entries for deferred revenue recognition.”

Full automation means that setting up the CRM meticulously is just a first step. The accounting touch-points cannot be afterthoughts: The users need to be trained to understand that the information in HubSpot will flow directly into the system. The added benefit here for CPAs is that if clients are meticulous about maintaining the CRM, the quality of the data will be high, which will improve the quality of the financial reporting.

Leverage Automation for Scalability

As finance professionals move from being stewards to strategists, everyone is looking for ways to grow. There are different strategies for increasing profitability in good times and bad, but one constant is the benefit of automating financial operations.

Because AI enables the processing and automated authorization of documents to enhance internal accounting processes such as procurement and purchasing, invoice processing, purchase orders, expense reports, accounts payable and receivables, you get more time to focus on the fundamentals — like financial performance. “Not everybody wants to do these manual tasks,” says Matt Merrill, Senior Manager at Tipalti, “but they keep the lights on.”

Once you have as much automated as possible, you’re in the best possible position for informed decision-making, whether it’s a choice to open a new office or pursue a new service line. “Make sure that your technology is aligned,” Matt says, “and you’ll be prepared for whatever swing we have over the next few years.”

How Generative AI Can Improve Internal Operations

Use ChatGPT for Initial Research

ChatGPT is a really good data-gathering resources. You can input a question and ask for results: “Can you bring me back a summary of how the accounting would go for this? How would you account for a lease or for new debt?”

That will be a huge time-saver when it comes to training new accountants.

The only catch: do you trust the resources the chatbot is using?

“Validate, validate, validate,” Kyle cautions. “Make sure the chatbot can point you to the resource the answer came from. If it came straight from the accounting codification, or a firm guide, great. If not, you need to be double-checking, especially with more technical areas.”

One recent development from OpenAI will help address this issue of quality sources: the possibility to create a custom GPT that only searches across certain files or websites, to filter out lower quality information.

Still, even if you do need to validate the sources, it still represents a significant savings of time. “That’s a lot quicker than leafing through all the research guides and getting your own answer.”

For people at smaller firms where there isn’t a guru sitting in the ivory tower, ready to dispense the answer, to be able to get that kind of start is huge.

“Build ChatGPT into your staff’s question process,” Kyle recommends.“ Tell them to come back with a proposed solution. It will take a lot of questions off the owner’s plate.”

Create an Internal Wiki to Share Knowledge

One of the major benefits of working with a team of accounting professionals, each of whom services multiple clients, is the volume of collective wisdom that results. But it can be time-consuming to figure out how to access all that knowledge: When a team member has a question and just throws it onto a Microsoft Teams’ channel, it might get lost.

Enter AI: Using note-taking software like FirefliesAI, OtterAI or Avoma to record meetings, it’s possible to tag the conversation, extract takeaways, and create to-do lists. But beyond the immediate action items, it also provides a search function.

Imagine a team member has a question about phantom stock plans. These AI note-takers will allow you to pull up the most recent internal conversations on the topic, allowing them to get up to speed — and to know who on the team is currently engaged in that issue and where to direct follow-up questions.

Just a few months ago, creating an internal wiki might have required a developer. Now it’s possible to get these capabilities at a really reasonable price.

Another way to use note-taker apps is for training videos: If someone makes a great point in an internal meeting, make a video cut, tag it, and share it with the team. While it might be hard to get busy team members to make time for an hour-long training, these video snippets are easy to consume.

These apps can also make it possible to save meeting time in general: If something important comes up, other team members can simply scan the transcript. They don’t need to sit for an hour in the meeting just to hear one important point.

How to Get Started with AI Accounting

With technology changing so quickly, it’s important for CPAs to dip their toes into things like ChatGPT and start getting some experience before the learning curve accelerates even faster.

“Have your ear to the ground whenever it comes to exploring new software,” says Adam Hale, Partner and Virtual CFO Practice leader at Summit CFO by Anders. “Make sure that you're, doing things like attending the dCPA where they talk a lot about this technology, or find some forums because it's less scary than what people think.”

Kyle has his own words of wisdom for the AI-hesitant, “As long as we're constantly exploring the new ideas that it can bring to the table, we'll be able to leverage it to improve our lives and make us more valuable to our clients.”