Beyond the Pilot: How CFOs Can Master the AI Value Chain

A: The key for CFOs is to start by clearly defining their own business problems or objectives, rather than simply adopting AI use cases that have worked for other organizations or chasing the latest technology trends. The most successful AI implementations in finance are those that are closely aligned with our organization’s unique priorities, needs, and circumstances.

What I advise is to look for a strong alignment between the scope of the use case and your own strategic goals, then evaluate each use case based on its anticipated value and feasibility. While many CFOs begin with areas like intelligent process automation or anomaly detection, we use frameworks like Gartner’s Finance AI and Opportunity Radar to help tailor these opportunities to fit our specific context and desired outcomes. Taking this deliberate, organization-focused approach helps avoid the underwhelming results that can come from more opportunistic or reactive use case selection.

A: Feasibility alone isn’t a proxy for value. A pilot may seem easy to implement due to organizational readiness or available skills, but it can still be low-value if it doesn’t deliver meaningful improvements. The clearest indicator is when a pilot merely automates existing work without materially enhancing the effectiveness of finance or its impact across the enterprise. When evaluating value, it’s important to consider alignment with strategic business goals, the total quantitative and qualitative benefits, how the use case fits within the broader finance portfolio, and the costs of experimentation, rollout, and ongoing operations. If a pilot fails to meet these criteria and offers limited benefits, it remains low-value regardless of how feasible it appears.

A: CFOs need to move beyond traditional finance skills and focus on building foundational AI and data science abilities in their teams, such as Python, data modeling, and prompt engineering. Developing “citizen data scientists” who combine technical skills with business acumen is key. Over time, finance teams should include more roles where humans and AI collaborate, allowing staff to focus on higher-value, AI-enabled tasks.

A: The best approach is to balance using embedded AI features in existing software for proven use cases (tactical AI) with building in-house AI for areas that offer competitive advantage (strategic AI). CFOs should act as “engagement architects” with vendors—ensuring vendors understand their processes and using innovation- or outcome-based metrics to evaluate performance. Starting with clear objectives and a hybrid “build and buy” strategy helps maximize value and reduce regret from AI investments.

A: To prevent AI costs from spiraling, CFOs should focus on creating shared accountability for AI spending across business units and key functions like finance and IT. This can be achieved through ongoing investment governance processes, such as AI venture boards that use a “Shark Tank”-style approach to vet and fund projects, or by adapting business cases into living project charters as AI is deployed. These methods help manage the main source of overruns — ongoing consumption and usage charges — rather than just initial build costs.

Additionally, CFOs should encourage cost-conscious behaviors by educating staff on the financial impact of AI usage, using incentives and finance-led training sessions. Ultimately, the most effective financial guardrails are outcome-based funding, milestone-driven fund releases, active governance of AI consumption and vendor contracts, and fostering shared accountability throughout the organization.

A: AI costs are volatile and can accelerate quickly, driven by usage-based pricing, changing vendor rates, and the need for strong data management. CFOs should take a proactive approach by closely scrutinizing cost drivers, demanding transparency from vendors —including hidden fees — and ensuring robust data governance. Since the value of AI is often unique to each organization, CFOs should also focus on value management by adopting a portfolio approach to AI investments. This means consistently measuring and discussing AI value across the C-suite, which helps compare use cases, evaluate business cases, and prioritize funding for the most promising pilots. Enterprises that use this approach are more likely to achieve mature AI adoption and deliver sustained ROI.