3. Surface the right questions. The most productive use of AI in complex financial analyses is not to arrive at conclusions right away, but to surface the right questions faster. Three approaches can make an analysis more effective:

Build a strong foundation. Tell the tool what you already know, what you don't know, and what could make the conclusion wrong. Break down a complex problem and feed the pieces into the AI tool. Incorporate the known risk areas, while understanding the existing data biases and limitations. It can also be helpful to ask the tool to reiterate its understanding of the task to ensure alignment.
Question the assumptions. Ask the AI tool to generate a list of key assumptions used in preparing the analysis, prioritize each of them based on the risk profiles, and validate against the actual source data. This is where professional skepticism matters most. An assumption may seem reasonable at first, but once you test it against the underlying facts, the analysis—or even the conclusion—may change.
Identify the blind spots. AI tools excel at data mining and pattern recognition. You can use them to flag potential blind spots in the financial data by asking questions like "What internal control gaps would allow this error to go undetected?" "What other areas could the lease accounting issue impact?" or "What are the early warning indicators, beyond payment history, for the current revenue trend?"

Consider the midsize software company with the revenue recognition challenge discussed earlier. Rather than asking AI to conclude whether the growth is real, the finance team could ask it to surface the questions a forensic reviewer would ask: "What is driving the concentration of this growth in the last month of the quarter? And if the growth is real, what should we expect accounts receivable and credit memo activities to look like in the next quarter?" Such questions would catch the problem, since AI would point the team toward the source data worth pulling rather than letting them accept a plausible-looking trend at face value.

Instant Insights /

What Forensic Accounting Can Teach Finance Leaders About AI

Related Stories

Recommended For You