In this conversation, Stephan Choe, Accounting Advisory Partner at WilliamsMarston, discusses how finance leaders can move quickly – but with discipline – to harness AI across the finance function, from automation and reporting to data governance, risk analysis, forecasting, and decision-making.

Q: Where should companies begin with AI strategy?

Many organizations feel pressure to “do something with AI.” That pressure is understandable, but if the underlying data is inconsistent, disconnected, poorly structured, or difficult to access, AI will not solve the problem. In some cases, it will make the problem more visible by accelerating the effects of poor inputs, weak processes, or governance gaps.

Companies should begin by assessing whether their finance function is actually ready for AI, including:

  • Quality of the organization’s data
  • Systems architecture
  • Scalability of current processes
  • Strength of risk and compliance controls
  • Role of human judgment 

Q: What does the flexibility that AI adoption adds mean for management teams practically?

Most companies don’t lack data; they lack direct, real-time, reliable access to usable information. Finance teams often spend too much time pulling reports, reconciling disconnected systems, cleaning data, or translating static outputs into actionable ones. AI can reduce that friction and give leaders a clearer view of what is driving performance.

That is where AI becomes meaningful for management teams, not as a standalone tool, but as a way to support faster, better-informed insights, including:

  • Stronger reporting
  • More efficient variance analysis
  • Better forecasting
  • Tighter alignment between financial and operational data

Q: How is WilliamsMarston thinking about AI in its own advisory work?

We see AI as both an evolution in advisory services and a strategic priority for clients. At WilliamsMarston, we are integrating AI into how we work, building new capabilities and investing in our teams while remaining focused on complex, high-value challenges where judgment, technical expertise, and close client partnership matter most.

Q: What AI-related issues should CFOs be watching most closely?

Two of the biggest emerging issues are visibility and agent sprawl. CFOs need to understand where AI is deployed, who is using it, how much it costs, and how it is being applied. As AI agents proliferate across functions, consistent governance will be critical to managing accountability, controls, and data integrity.

Q: How is AI changing the way companies think about ERP systems and finance architecture?

For decades, companies have relied on large, one-size-fits-all ERP systems, often paying for capabilities they do not use and adapting their operations to fit the technology – rather than the other way around. AI is accelerating a shift toward more flexible architectures designed around how the business actually operates.

Q: How should companies evaluate whether their current ERP systems are helping or holding them back?

Finance leaders should recognize where their systems add flexibility or friction. Frequent exporting, reconciling, manual cleaning, or workarounds indicate architecture lags behind business needs. AI won’t fix this alone and may reveal these weaknesses more clearly.

In practice, that means CFOs should be asking fundamental questions, such as:

  • Can we trust our financial data?
  • Can our systems support how the business actually operates?
  • Are our workflows flexible enough to adapt?
  • Do we have the controls needed to use AI responsibly?
  • Are we solving a real business problem, or are we reacting to market pressure?

That is where AI becomes valuable: when it is embedded into real workflows, tied to real business questions, and supported by the right data structure.

Q: What should CFOs and finance leaders anticipate next?

The first step is to make AI a business conversation, not just a technology conversation. Beyond costs and individual tools, CFOs should be working with their teams to evaluate:

  • The organization’s data foundation
  • Systems architecture and process maturity
  • Where AI can create measurable value
  • Where additional governance, controls, or change management may be needed
  • Talent training and tool needs

AI will continue to evolve quickly. But the underlying principle is straightforward: companies need clean, usable data, flexible systems, and clear governance before AI can become a real advantage.