Across the finance organizations we have worked with, we have found that interest in AI is real, but practical adoption is uneven. The clearest benefits come from bounded tasks with a named owner and a human checkpoint. Where the opportunity is unrealized, the barriers are familiar: data gaps, system transitions, stretched teams, and an unclear business case.
What We Are Seeing
Usage is personal, not institutional. People use AI to draft a memo or clean a data set, but many enterprise licenses sit unused.
A leader who clears the way wins. Teams move fastest where a CFO or controller sets the expectation and protects time to build.
Show me first. Then they want it yesterday. People do not believe the time savings until they see the result, and then they want it right away.
The human stays in the loop. AI drafts, and a capable human still owns the judgment.
Builders vs. the hand-it-to-me crowd. A few people lean all the way in, but most just want a finished tool. The biggest gains come from problem engineers, not those solely focused on automation and execution.
The tool landscape is fragmenting. Copilot, Gemini, ChatGPT, Claude, and model-routing aggregators all show up, and CIOs are avoiding lock-in.
The static dashboard is fading. Interactive reporting that used to take weeks is now mocked up in hours.
Where Client Adoption Stands
We see finance teams in four places, often within the same organization. Moving up depends less on the tools than on data, ownership, and controls. Each step earns the right to the next one.
What Makes AI Value Last in Finance
Traceability Beats Speed
Reviewers need a trail: formulas, visible steps, and week-over-week bridges. Untraceable, hard-coded AI output adds rework, and variance commentary still needs someone who knows the business.
Ownership Decides What Lasts
Building the tool is the easy part. At one client, a searchable contract database stalled over legal and IT expectations and who would maintain it. Agree early who benefits, who maintains it, and who carries the risk.
Strong Foundations Come First
AI amplifies whatever it sits on. Clean data, sound processes, and clear controls matter more than the tool, and so do people who can catch a wrong answer instead of trusting every result.
Access Is Not Adoption
Fix the finance process first, then add the right tools. Show teams a working example using their own data, give the tool a willing owner, and involve IT early. Build in training and technical support. People need the time and skills to use the tool well, not just access to it.
Where to Start
Pick one process you run every month. Close, the reporting pack, or reconciliations. Time it today so you can prove the gain later.
Name an owner before you build. If no one in finance will maintain it, do not start.
Keep AI on draft. A named person approves every entry, and every number traces back to its source.
Price the run, not just the build. Licenses, usage, and upkeep belong in the business case from Day One.
Make the first demo use their numbers. A working example on their own data beats any training deck.
What It Means for Finance Leaders
AI is not a finance strategy on its own. Nor is it a software that you simply plug in. The organizations making progress treat it as one tool for a well-run finance function: a defined problem, reliable data, a named owner, and explicit approval. Prove one step, then expand when the economics are clear.
© Copyright 2026. The views expressed herein are those of the author(s) and not necessarily the views of Ankura Consulting Group, LLC, its management, its subsidiaries, its affiliates, or its other professionals. Ankura is not a law firm and cannot provide legal advice.
