Financial leaders around the world are feeling increasing pressure to implement artificial intelligence. But they struggle with the question of how to actually derive value from it. This is clear from a new study by Busware.
According to the report “AI to ROI: Unlocking Value with AI Agents,” nearly half of CFOs say board members and executives are increasingly demanding AI solutions within finance processes. At the same time, the majority indicate that their organizations are primarily experimenting with AI, without a clear picture of real-world applications or return on investment. For example, 61% say they primarily experiment with homegrown AI agents to see what the technology can do. A quarter of respondents even admitted that they still don’t fully understand what an AI agent actually looks like.
agent AI
The emergence of so-called “agent AI” – systems that can independently perform tasks and support decision-making – is playing a central role in this. Two-thirds of respondents believe there is currently more hype around agent AI than previous technological developments. At the same time, three-quarters have yet to discover how to use this technology effectively. According to Basware CEO Jason Kurtz, the time for experimentation is over. Management expects concrete results.
The numbers show that organizations that use AI in a targeted manner are achieving better results. The average return on AI investments rose to 67 percent from 35 percent last year. Companies using agent AI via existing platforms with built-in AI agents achieve an average return of around 80%.
Data-driven components of the finance function
For many finance teams, the first application of AI is in the accounts payable process, which is often the most manual and data-driven part of the finance function. In fact, 72% of respondents cited this as the most logical starting point. Other applications include invoice automation, cash flow management, forecasting, compliance checking, and fraud detection.
The key, Basware says, is to use trusted data and integration platforms rather than standalone solutions or home-built systems. With solutions such as InvoiceAI, part of its invoice lifecycle management platform, the company aims to improve speed, accuracy, and compliance while using AI agents to handle invoice processes almost autonomously.
the study
The study also shows that strategy and governance matter. Teams with the lowest AI returns often report acting under pressure without clear direction. Organizations with a clear approach achieve better results.
The report’s conclusions are clear. AI can significantly improve financial processes, but only if organizations deploy the technology in a targeted manner and integrate it into existing workflows. Only then will AI move from experiment to real business value.
Also read: Goodbye AIOps, hello AgenticOps: What is it? What can you do with it?
