Finance departments have traditionally been risk-averse, which has often led them to lag in adopting new technologies. This writer recalls finance leaders insisting that their company’s financial data was too proprietary to ever move into the cloud. Yet, this caution hasn’t always been the norm. Finance was among the earliest adopters of personal computers. PC-based spreadsheets revolutionized how financial work was done, transforming processes once handled on paper with a Texas Instruments or HP calculator. Those manual methods were slow and error-prone, so it was a godsend when spreadsheets made financial analysis faster, easier and far more accurate. 

In fact, PCs became a status symbol in accounting — public accounting firms proudly showed off that everyone had the latest PC. Geoffrey A. Moore, in “Crossing the Chasm,” writes about the role of Lotus 1-2-3 in enabling its delighted early adopters “to do something they had never been able to do before — what later became popularized as ‘what if’ analysis.” 

The question now is whether generative and agentic AI will fundamentally reconfigure how finance is done. Bruce Harris, director of financial systems and intelligence at Torchy’s Tacos, put it well in a recent interview with me. 

Related:Building an MCP server is easy, but getting it to work is a lot harder

“Every taco we sell is in our cloud data warehouse, and this data tells a story. By embracing agentic AI, we’re transforming finance from transactional to strategic,” he said. “Our agentic workflows automate the routine work, freeing our people to focus on insight, strategy, and growth. This isn’t about replacing talent — it’s about amplifying it.” 

To explore this shift further, I spoke with experts at three ERP companies that are enabling agents for their finance customers:

Their perspectives are surprising and deserving of wider attention — especially for CIOs, I would wager. For many organizations, CFOs have been the executives to whom IT reported — or, at minimum, one of IT’s most demanding and consequential internal customers. Countless CIOs have seen their lives upended by ERP implementations that dragged on for years, consuming budgets, attention and every available set of hands. These “all-hands” moments have repeatedly locked CIOs into long cycles of implementation and reimplementation. 

What will be interesting to watch now is whether the shifts underway — particularly, finance’s push to apply AI to become leaner, more automated and more strategic — trigger another implementation cycle. Interestingly, if finance can reimagine its operating model, CIOs may find themselves at the center of a very different partnership with the CFO.

Related:Hot chips, cold feet: What happens when AI’s infrastructure outpaces demand?

From number crunchers to strategic advisors

Each of the ERP experts I spoke with made it clear that AI agents will automate transactional and compliance work, freeing finance professionals from manual tasks, including data entry, financial reconciliation and expense validation. With agentic AI, the boring, repetitive financial work is officially over — a welcome development for someone who had done financial analysis right after my first MBA. It was not my calling, but people who were STs in a Myers-Briggs assessment thrived in traditional accounting-type roles. What will this mean for those types? 

AI agents will refine accounting and finance roles from transaction-heavy to insight-driven, shifting focus toward strategic analysis, decision support and business partnership. The hope, clearly, is that with the support of AI, finance teams can tackle previously “undone” work, unlock new productivity and enable faster, smarter business decisions.

Related:Make your own mandate: How CISOs can implement GenAI governance

The AI opportunity for the CFO role

Here are excerpts from my discussions with Kershaw, Preston and Alavarez (lightly edited for clarity and brevity) on the importance and implications of applying AI to finance, starting with how AI will redefine the role of the CFO. 

Andrew Kershaw, Workday: “Agents will accelerate the evolution of the CFO’s role, enabling [them to spend] the vast majority of their time on strategic opportunities across the business vs. managing transactional efficiency within their group. The core goal has always been the same: less time on transactions, more time on insights that drive the business forward. The value of agents lies in automating finance processes to help the CFOs and their teams both protect and grow value in the business. 

“On the protection side, it’s about automating for greater accuracy, compliance and risk mitigation. On the growth side, it’s about unlocking insights to drive the business forward. By taking on tedious work that doesn’t require human judgment, agents free up teams to focus on strategy and high-value decisions. … This is how CFOs gain the credibility and capacity to stop spending time looking back and start spending it looking forward.

“It’s exciting because agents are moving beyond just surfacing insights to actually taking autonomous action, delving deeper into the data to understand variance or root cause of issues, then resolving an error or notifying the right people — effectively automating the workflow from insight to resolution.”

Joe Preston, Intuit: “While most financial tools give CFOs access to data … it’s challenging to cut through the noise and determine what’s valuable. Agentic AI identifies trends, connects and finds insights that are overlooked or hidden, helping CFOs understand not only where their business stands today but where it’s headed. Agents can provide a comprehensive approach to the financial management of growing, midmarket businesses with robust reporting, KPI analysis, and scenario planning and forecasting based on performance and peer benchmarking, helping CFOs and their finance teams make smart decisions to achieve their goals.”

Victor Alvarez, product marketing manager for Joule, SAP

A new division of labor in finance 

AI agents are expanding automation by handling complex, multi-step and cross-functional workflows like invoice matching, cash collection and dispute resolution — while improving speed, accuracy and cash flow. With this said, our ERP experts noted that human expertise remains central. 

Kershaw: “What sets AI agents apart is their ability to automate parts of finance that couldn’t be automated before. Past solutions struggled with ‘gray areas’ — tasks requiring judgment or cross-functional input. Now, agents handle these complex, insight-driven tasks, making finance workflows smoother and smarter. For example, in accounts payable, if an invoice doesn’t match a closed purchase order, agents can handle this autonomously, coordinating with other agents to resolve the issue, while still respecting the control environment. 

“However, while agents are great at surfacing data and routing decisions, human judgment remains critical, especially for complex financial decisions. Agents will make it easier for decision-makers to act with confidence, but decisions that impact financial results require human oversight because someone needs to own the outcome. For example, AI can surface data for bonus accruals, but leadership must make the final call because executive alignment is required.”

Victor Alvarez, SAP: “Agents will handle common, multi-step workflows that require reasoning over data and business process context (e.g., invoice processing, dispute resolution, trade classification). They’ll also perform cross-functional workflows, such as cash collection involving finance, customer service and operations. Real-time decision support through recommending actions based on trusted, high-quality financial data is another significant benefit. For example, an accounts receivable agent doesn’t just automate receivables. It reasons through open items, balances, disputes, and dunning history to assess risk and prioritize follow-ups. It analyzes this context to flag high-risk receivables, recommends the next best actions and guides users with proactive, timely insights. Then it acts — initiating follow-ups, prompting responses and supporting resolution. This can result in less time spent managing overdue receivables, fewer write-offs through early risk detection and improvement in DSO to strengthen cash flow.”

Joe Preston, vice president of product and design, Intuit

Can finance learn to trust AI with its data?

AI complements — does not replace — human expertise, with people providing essential context, oversight, and ethical judgment in decision-making. Security, data integrity and privacy are paramount but will require finance leaders to understand how AI reaches conclusions to ensure accountability and compliance. 

Kershaw: “Beyond the need for AI to act in an auditable, correct and repeatable manner, currently, the biggest hurdle for finance organizations isn’t understanding the value of AI — it’s reimagining what’s possible and adopting new ways of working. On reimagining possibilities, finance leaders aren’t used to AI agents providing instant, strategic recommendations instead of their having to manually track down information. Regarding new ways of working, finance teams must adapt to new workflows, including closer collaboration with IT.”

Preston: “Organizations need to keep in mind that AI complements human intelligence. While AI automates certain tasks and surfaces valuable information, human expertise is critical to ensure the right context and decision-making is applied. It’s also important for firms to remember that public AI tools may lack the secure environment needed when analyzing client data.”

Most exciting tasks to automate with AI agents?

The biggest challenge for finance leaders is not about recognizing AI’s value but reimagining what’s possible with AI. Here’s Kershaw’s take. 

Kershaw: “Two areas: contracts and cost/profitability analysis. They are exciting because they represent the removal of very time-consuming and cumbersome activities that unlock incredible value. 

“First, consider contracts. With a revenue contract agent, for example, AI automatically reads incoming contracts, sorts them by type and extracts all the critical data points like customer name, payment terms and total contract value. Crucially, the AI is continuously monitoring your entire portfolio and surfacing key insights through interactive dashboards, giving finance professionals insight into things like built-in rate increases tied to inflation that could automatically expand your revenue. 

“Second, profitability isn’t just about one big number; it requires analyzing the true cost of operations for both direct and indirect costs and providing clear transparency into how shared resources are consumed. Agentic AI allows accountants and finance professionals to allocate indirect costs daily — such as management fees, utilities, IT and marketing — down to the individual outlet.”

Parting words

As each of the ERP experts made clear, finance organizations are on the precipice of significant change. For a profession developed as Columbus sailed the ocean blue, the change and disruption that it is about to experience is earth-shattering. In “Epic Disruptions: 11 Innovations That Shaped our Modern World,” Scott D. Anthony writes that “disruption is an engine of progress. By making the complicated simple and the expensive affordable, it transforms how we work, play, live and communicate.” Nowhere will this transformation be clearer than in accounting and finance as agentic AI takes hold. 

In a world where the books of the company largely run themselves, it will be the more cerebral accounting and finance people who are in demand. These survivors will not only understand the books but also be able to make concrete suggestions on achieving business transformation. 

Demonstrating this line of sight into business transformation will be a challenge similar to what happened to the CIO and their teams since the COVID-19 pandemic: the ones who survived underwent personal transformation, in many cases adopting a new mindset and skill set. 

This time, the personal transformation is required by the CFO and their key reports in order to lead the next wave of change. And just like with CIOs and their teams in the wake of the pandemic, not everyone will be capable of making the change.