• Overall, Finance is making an increasingly important contribution to effective steering
• Routine processes such as accounting entries and invoice processing are already fully automated in many cases, while CFOs are taking on greater responsibility for value creation, data governance, and process management
• Traditional finance expertise is no longer enough: Finance teams need new skills in AI and data

Artificial intelligence (AI) will fundamentally transform corporate finance functions: Routine financial accounting processes such as accounting entries and invoice processing will be automated in the future and already are at leading companies today. As standardized day-to-day tasks are eliminated, the cost of the finance function will fall from 0.8 percent of revenue to less than 0.5 percent over the next five years. At the same time, the CFO's role is growing in importance: The CFO will become a key driver of value creation within the company. These are the findings of the Horváth analysis “AI-Based Finance TOM.”

The reason for this development is the increasing complexity facing finance executives. Markets, supply chains, and regulatory requirements are changing more quickly, while at the same time, more and increasingly detailed information is required as a basis for decision-making. In eight theses, Horváth experts describe how CFOs can meet these new requirements, from reallocating today’s tasks between people and AI agents to restructuring IT systems and developing new skills within finance teams. As a result, the finance function’s understanding of its own role is changing: “In our projects, we see that companies initially view AI purely as an efficiency topic. They use the technology as a tool to handle existing processes faster, but that does not go far enough. The actual shift will take place at the level of responsibility: The CFO will become the architect of business steering. Those who focus only on costs will miss the real opportunity,” says Stefan Tobias, Partner at Horváth.

New Division of Roles Between People and AI

In practice, this shift is already becoming apparent in day-to-day operations. Processes such as invoice processing, account reconciliations, and closing entries are increasingly being performed automatically and without human intervention according to a “Dark by Default” principle. Human involvement will, however, remain necessary wherever decisions carry high risks or irreversible consequences, such as capital allocation, M&A decisions, or the approval of large credit limits. This is changing how traditional CFO responsibilities are divided between people and AI agents. Particularly in operational finance processes, AI agents and bots are expected to take over the majority of execution activities, while employees focus on critical exceptions and the further development of governance. The extent to which this division will shift within individual companies depends on their specific circumstances and must be assessed on a company-by-company basis.

Finance Teams Need Different Skills and a New Systems Landscape

The changing division of roles has an immediate impact on the capabilities required of Finance teams. While traditional expertise in accounting, controlling, treasury, and tax will remain fundamental, new areas of expertise are coming into focus. These include the ability to work confidently with AI systems as well as data skills. For companies, this means a greater need for hybrid finance-tech profiles, continuous upskilling of existing teams, and closer integration between Finance, IT, and data management.

These new capabilities coincide with an IT landscape that will also undergo a fundamental transformation. Today’s widely used system architectures will be replaced by cloud-based platforms on which AI agents independently execute processes and proactively prepare decisions. A practical example already shows how the benefits of such systems can be measured: A controlling chatbot achieved a user satisfaction score of 8.5 out of 10 and a response accuracy of 95 percent. As a result, ten controllers saved 2.5 hours of working time per month.

As AI agents become increasingly important, the need for consistent data standards is also growing. Today, responsibility for data generally lies with the respective functions: financial data with the CFO, sales data with the head of Sales, and employee data with HR. This fragmentation makes it more difficult to establish a reliable basis for AI-supported decisions. Going forward, the CFO will therefore set governance standards across domains. “Only a clean, company-wide data foundation makes AI-supported decisions truly trustworthy. That is why this responsibility should lie with the CFO, as the company’s business conscience. The CFO already has the discipline and systematic approach needed to ensure reliable data. This makes him the natural anchor point for governance across the entire organization,” says Stefan Tobias.

About the Analysis

The Horváth Point of View “AI-Based Finance TOM” sets out eight theses describing how artificial intelligence will transform corporate finance functions in the years ahead. The analysis draws on Horváth's project experience and was published in September 2026.