As last month's AI for Finance Forum in New York City (co-hosted by AWS and CFO Leadership) drew to a close, the closing keynote uttered an unpleasant analogy: “Finance is like a duck.” On the water, he is calm, collected and delivers on time. Behind the scenes, you're frantically working on manual adjustments, spreadsheet gymnastics, and low-value tasks that take up the bulk of your team's time.
After spending a full day learning about AI partners and their use cases, the room full of financial industry leaders knew exactly what was going to happen next. The AI gives you the chance to stop paddling and start swimming. But this metaphor stuck because it reframed everything we heard that day around the only question that actually mattered: how do we move from AI pilot to real business value?
6 lessons that will shape the future of finance
1. Technological literacy is no longer optional
You don't have to be a data scientist, but you do need to understand that AI is about probability, not magic. AI models make predictions based on patterns in your data. they could be wrong. These reflect the quality of the data and context you provide.
This understanding changes the way you evaluate AI solutions and set expectations for your teams. It helps you ask better questions of your vendors and partners. Most importantly, it helps guide your organization through change with authenticity. For financial leaders, literacy means more than just knowing that a model is probabilistic. Step into co-ownership of the AI stack and partner with the CIO and CDO on which use cases to prioritize, which data is “authoritative,” and what level of risk is acceptable for which decisions.
2. AI will continue to evolve
Today's AI is the worst ever. The pace of progress means that models will become more capable, more accurate, and more accessible over the next 12 to 24 months.
This has two implications. First, don't expect perfection. Start learning now, even if the tools aren't ideal. Next, build flexible systems and processes. How you currently implement AI may need to evolve as capabilities improve. Don't build a single “AI project”. Build a living AI roadmap. Expect models, tools, and even agent responsibilities to change every 6-12 months. Winning finance teams will be those that can quickly adapt their workflows as technology evolves.
3. Start with the problem, not the technology
Too many organizations are asking, “Where can I use AI?” A better question is, “What problems are I spending the most time, money, or risk on?” One forum participant described how her team spent many hours each month coordinating intercompany transactions across many subsidiaries. This was an error-prone and daunting task.
Where is your team spending time on repetitive tasks? Where are errors causing downstream problems? Where are they unable to scale because their processes don't allow them? Start by mapping those pain points. Then assess whether AI can deal with them.
4. Measure value, not activity.
Your AI strategy should answer one question: What business outcomes will it improve?
Define success before you start. Will this reduce your cycle time? Free up your time for higher-value work? Reduce errors that create financial risk? Do you want to improve the quality of insights you provide to your business partners?
The most successful AI implementations have clear metrics tied to business value. Struggling teams can’t articulate what success looks like beyond “I’m using AI.”
5. Create space for experimentation
Innovation requires competence. Teams can't experiment with new approaches if they're buried in spreadsheets and manual processes.
Forum example: A finance leader asked team members to forego Excel for certain analyses, and instead use new AI-powered tools. This forced experimentation and created space for teams to learn, make mistakes, and discover better ways of working.
AI not only speeds up existing work, but also enables contributions that were previously impossible. Team members who had ideas but lacked the technical expertise can now turn those ideas into reality. Analysts who have never learned Python can now build custom tools. FP&A managers with a hypothesis can now test it without waiting for data science support.
If your team is at 110%, the AI becomes another thing you don't have time for. You need to make space by dedicating dedicated study time, pausing non-essential initiatives, and reducing meeting load to free up capacity.
6. Business context becomes a competitive advantage
Every company has access to the same AI models. What differentiates the results is the business context they provide.
Document processes, capture organizational knowledge, and make tacit expertise explicit. When an AI agent understands chart of accounts structures, revenue recognition policies, and business unit nuances, it delivers better results than typical implementations.
The business context transforms AI from a commodity to an asset.
move forward
The successful financial leaders over the next five years will not be the ones with the most AI projects. They will use AI to solve real problems, deliver measurable value, and build technical literacy across their teams to do this more efficiently.
Start with one high-impact problem. Build small teams with protected capacity. Define clear success metrics. Document your business background. Experiment, measure, learn, and iterate.
The question is not whether AI will transform finance. It's whether your team will still drown in spreadsheets while your competitors are already swimming laps.
