The Smart CPA Firm: Building a Successful AI Ecosystem
As AI moves from possibility to business necessity, many accounting firms risk rushing in without thoughtful adoption and integration.
By Teri Saylor | Winter 2025
The growing importance of artificial intelligence (AI) in the modern workplace is reflected in recent data. A 2025 Thomson Reuters Future of Professionals survey of 2,275 professionals and C-suite leaders revealed that 80% of respondents believe AI will
have a high or even transformative impact on their profession within five years, and over half believe their organization is already benefiting from the use of AI.
Yet, despite growing enthusiasm, many organizations, including certified public accounting
(CPA) firms, are realizing that successful AI adoption involves more than simply adding new tools—it requires thoughtful and meaningful integration with existing systems, processes, and people, and navigating this transition requires strategies
for adopting and embedding AI in ways that strengthen, rather than disrupt, current technology ecosystems.
Getting Started
Building an AI ecosystem isn’t just about adopting next-level technology—it’s about building a data-driven environment
that allows intelligent systems, people, and processes to work together.
David Fuge, chief information officer in the Virginia office of Johnson Lambert LLP, recommends beginners simply start by interacting with the AI tools already accessible to them
and becoming fluent in them.
“I think people just need to jump in and start getting their feet wet while building a sense of literacy around how to use AI effectively,” Fuge says. “How do you communicate with it, what happens when you
ask it for certain things in certain ways, and are you getting what you want from it?”
Megan Angle, CPA, a partner at Porte Brown in Chicago, recommends experimenting with a single undertaking first. For example, try integrating generative AI into
your normal processes and aim to reduce the amount of time it typically takes to generate and articulate ideas and complete tasks. “Then use AI to track results, including time savings per engagement, turnaround time for deliverables, and error
reduction,” she suggests.
As your AI comfort grows, Angle recommends launching a pilot project in areas with high transaction volume and intense manual labor, focusing on repetitive, rules-based tasks, like bank feed classification, tax document
extraction, or engagement letter creation.
Using popular apps, like ChatGPT, Microsoft 365 Copilot, Claude, and Gemini, is good to start with, Angle says, but also experiment with the AI features in software you may also be familiar with, like QuickBooks
and Sage. Another option is to explore modular AI applications, like Dext, Docyt, or DataSnipper, which can connect to existing systems.
“Our firm uses DataSnipper, which integrates with Excel to automate repetitive tasks like data extraction and
cross-referencing from source documents,” Angle explains. “It uses AI and optical character recognition to extract text from PDFs, documents, and images that we scan into our system.”
For Ayala Clinkman, CPA, PMP, senior director of
business solutions at Business Technology Partners in Chicago, ChatGPT is the go-to AI tool for tracking time. Using it, she can analyze timesheets in minutes and spot areas where she can improve efficiency and boost the bottom line. She also uses ChatGPT
to analyze financial reports and large volumes of data for spotting trends that could help her firm manage its budget and streamline practice management processes.
ChatGPT has even helped Clinkman improve client communication. She recalls a client who
emailed her requesting a fee reduction: “I was able to provide ChatGPT with all the nonconfidential information I needed to convey to my client, and it helped me answer their questions in one, comprehensible message that didn’t come off as
rude or groveling.”
The Necessity of Safeguarding Data
According to Randy Johnston, executive vice president of K2 Enterprises, privacy, compliance, and transparency aren’t optional in the world of accounting, but users generally don’t
realize the vulnerability of data shared or stored in AI tools.
“So many users don’t understand the security and privacy implications of the AI engines, and they’re inputting way too much personally identifiable data and bank and financial
statements into them,” Johnston cautions. “Users have the potential to expose confidential client information, open channels for cybersecurity risks, and facilitate unintentional data leakage.”
Notably, misuse of AI tops the list of
concerns among practitioners who completed the Thomson Reuters survey, with 23% of respondents reporting concern over unethical use of AI.
However, good policies and a strong governance framework can help convert AI from a risk factor into a strategic
advantage, ensuring that automation enhances human expertise rather than undermines it.
In the summer 2025 Insight article, “6 Keys to AI Adoption in Accounting and Finance,” Fuge and his colleague Andrea Wright, CPA, identified key elements
that should be outlined in an AI governance strategy. They noted that policies should have clear answers and controls around:
- The data that should be collected and stored.
- How data is processed and used.
- Who has access (and under what
conditions).
- How data accuracy, completeness, and consistency are maintained.
- Sensitive and confidential client data is protected.
- How long data is retained and securely disposed of.
- How compliance is ensured.
For Clinkman, any
use of AI must be vetted: “Any platform or tool needs to be approved and deemed safe and secure. To avoid the risk of disclosing client information, we don’t even use transcribing tools.”
Likewise, Fuge advises avoiding using the free
versions of AI tools and carefully vetting any software your firm subscribes to.
“If you’re going to use AI tools and input client data into them, you need to make sure you’re doing so with appropriate confidentiality agreements in place
and a substantial third-party risk management vetting process,” Fuge says. “It’s important to include guarantees around data retention and what the software providers can and can’t do with the data.”
Fuge outlines these safeguards
for protecting data:
- Never upload sensitive data to free AI platforms.
- Ensure the appropriate confidentiality agreements are in place with third-party vendors.
- Clarify data retention rules and how long the vendor may store your information.
- Maintain controls to prevent data breaches.
- Establish strong policies around password controls and the use of AI on personal and mobile devices.
Selecting AI Software in an Overwhelming Market
The AI marketplace is being flooded with tools spurred by rapid innovation, new integrations, features, and capabilities.
Johnston, who often consults with firms on technology needs and vendor contracts, recommends starting with a single reliable platform that ensures trusted systems will operate as intended—safely, securely, and free from harmful consequences. He recommends Claude and Microsoft 365 Copilot.
Additionally, when it’s time to expand, Johnston advises taking a “less is more approach.”
“A few years ago, I started cautioning people against having too many subscriptions, applications, and integrations, which all lead to too much complexity,” he stresses. “Don’t just throw a lot of technical tools against an issue. Instead, be more intentional and strategic with the tools you already have.”
“Technology is moving so fast, you can find yourself checking the marketplace every six months to explore what’s new,” Angle acknowledges, which is leading many firms to begin internally developing AI tools to help cut through the noise of the crowded marketplace and find a solution that works best for their needs.
“There are a decent number of firms that have hired data scientists and people who are writing code and creating their own AI components,” she says. “While not feasible for every firm, it’s exciting to see what these firms are doing and what they’re capable of.”
Johnston gives a nod to the growth in both volume and complexity of software and how difficult it can be for firm management to define what they need, especially if their practice includes tax, audit, and client accounting services. Still, defining needs as clear as possible is essential before venturing into the software marketplace or developing an internal solution—AI decisions need to be based on facts rather than emotions.
Overall, when considering purchasing or developing an AI solution:
- Start by mastering a single platform, keeping safety and security in mind.
- When expanding, take a targeted approach and avoid subscribing to unnecessary software.
- Do a needs assessment based on your firm’s services and business practices.
- Consider a system that fits your firm’s size, needs, and culture.
Fostering AI Literacy
The Thomson Reuters survey found that organizational leaders who proactively reconfigure their workforces for the AI era will gain a competitive advantage by aligning talent with technology.
At Porte Brown, Angle says implementing new ideas is spread across multiple roles: “We have a director of quality control, a team that champions new concepts and ideas, and a group of partners who are constantly reviewing and evaluating AI implementation within the firm.”
She believes the use of AI is leading to a paradigm shift that’ll change job descriptions for most accounting professionals. She predicts it’ll eliminate 90% of their hands-on tasks and free them up to spend more time on higher-value work, such as analysis, advising clients, and interpreting results rather than just gathering them.
“It’s just a matter of how we’re changing our staff development initiatives to quickly get them to where we need them to be,” she says.
Fuge and his team are leveling up their younger staff by leveraging AI tools to enable them to complete more complex tasks.
“There’s a lot of buzz around AI taking accounting jobs and eliminating junior roles,” Fuge says. “However, we see this as an opportunity to enable junior staff to do what used to be reserved for our more senior professionals.”
Johnston notes that many CPA firms are going through this change cycle right now. To better position themselves for this shift, Johnston says firms should evaluate their current staffing model and invest in training and upskilling (which can be done internally).
“When it comes to AI and training, we advocate for taking a continuous learning approach—investing in internal training is almost mandatory,” Johnston says. “Training programs work best when small teams attend educational events to get new ideas, then bring them back and teach them to everybody else.”
To build an effective AI-ready team:
- Determine staff roles, such as quality control lead and technology champions, that fit your practice.
- Implement the use of AI to level up early-career staff, enabling them to complete more complex tasks.
- Focus on internal training and upskilling.
- Encourage staff to take a self-directed approach to learning through experimentation while keeping guardrails in place.
- Establish open communication channels for sharing ideas, successes, and failures.
The Time Is Now
No doubt, the hesitation around AI is real: It’s new, it’s fast-changing, and it can feel overwhelming for many organizations. But the experts emphasize that waiting for the right moment to adopt it is no longer realistic.
“We’re seeing AI integrated into every program we touch, and we’re going to go there whether we like it or not,” Angle stresses. “It’s not even about figuring out when we’re ready for it because we’re already there.”
Johnston assures fledgling users to never fear AI, reminding them it’s just a tool to make our jobs easier: “Using AI is all about creating task-specific, role-based technology that automates workflows and saves time.”
Fuge stresses it’s really just about getting started and comfortable with it: “Using AI regularly and looking at it like just another managerial tool is a good way to get started. Now’s the time to hone your skills so you don’t fall behind.”
Teri Saylor is a freelance journalist in Raleigh, N.C. who writes about business and lifestyles.
Related Content: