Finance departments have spent decades as the organizational gatekeeper, the function that other teams wait on for analysis, approval, and answers. That model is breaking down, and the chief financial officers (CFOs) who understand why are the ones pulling ahead. Dr. Mitchell T. Horn, a finance and financial planning & analysis (FP&A) leader with deep experience in consumer packaged goods (CPG) organizations, has spent the last several years building AI into the core of how finance operates, not as a productivity tool bolted onto existing processes, but as a structural shift in what finance is for and who can access its outputs. “People are enjoying the independence and the information they can gather very quickly and very accurately,” Horn reflects. “They’re making decisions faster than they ever have.”
Finance’s New Role Is Validation, Not Bottleneck
The most significant change Horn has driven is deceptively simple: giving business leaders, marketing, sales, and strategy direct access to AI tools capable of performing the financial analysis they need to make decisions. Teams no longer wait for an email from finance or schedule a meeting to get a business case evaluated. They run the analysis themselves, in real time, and finance validates the output.
That validation role is not a demotion. It is a repositioning toward higher-value work. Finance leaders bring the reference material, contextual expertise, and judgment to confirm whether AI output is decision-grade, a level of domain knowledge the AI cannot replicate independently. The sales and operations planning (S&OP) process at Horn’s organization illustrates the operational upside concretely: the AI tool Claude significantly streamlined forecasting, removing the debate and rework cycles that had previously consumed planning bandwidth. The process used to generate friction now operates with considerably less friction.
The implementation path mattered. Horn’s team started with ChatGPT before pivoting to Claude, which drove near-immediate cross-functional adoption. The reason was straightforward: the output was reliable and accurate enough that confidence followed naturally. The evolution continued from web-based access to the desktop application, and then to Claude add-ins for Microsoft Excel. That last step, Horn describes as the breakthrough moment for the finance team, the point at which AI moved from a supplementary tool into part of the financial model itself.
The AI-Enabled Finance Organization Is Already Here
Horn’s outlook on the next two to three years for CFOs and finance leaders is direct and carries real stakes. His organization has a hiring freeze for finance talent, and the expectation when someone exits is not backfill; it is absorption by remaining staff, supported by technology that takes on the work the departing headcount was responsible for. Finance organizations are going to become significantly leaner. Fewer hierarchical levels will be required. A technically capable CFO, in Horn’s view, eliminates the need for a substantial portion of middle management.
Horn is already operating three finance Agentic AI agents and consistently updating their capabilities. Two challenges remain unsolved: controlling what information agents can access to protect sensitive data and calibrating how finance agents communicate with non-finance users who approach them with different expectations. The CFOs who are excited about where this is heading are the ones upskilling consistently. The ones who are not, by Horn’s account, are deeply uncomfortable about the future of their careers, and that discomfort is well-founded. “It comes down to the individual,” he states plainly. “Either aggressively upskill and invest the time to stay relevant, or be replaced by someone who has.”
Follow Dr. Mitchell T. Horn on LinkedIn or visit WM Partners, LP, for more insights on AI in finance, FP&A transformation, and building the technically capable finance functions that lead organizations rather than serve them.
