AI took over Business has been in the headlines for the past two years, and finance is no exception. In an exclusive McKinsey survey of 102 CFOs across industries and global regions, 44% of respondents said they would use Gen AI in five or more use cases in 2025, up from 7% in the previous year’s survey. Investment in AI tools is also increasing. 65% of respondents said their companies will increase their AI investments in 2025. Two years ago, only about a quarter of respondents said the same.
But the reality across the enterprise highlights how elusive tangible value can be. In a recent McKinsey survey, nearly two-thirds of respondents said their organizations have not yet begun scaling AI across the enterprise. Poor results are primarily due to pilots being dysfunctional under real-world conditions, unable to adapt to new data as it emerges, and poorly integrated into core processes.
But some finance teams are successfully leveraging AI, generative AI, and, increasingly, agent AI to increase efficiency, improve insights, and offload time-consuming manual tasks (see sidebar: A guide to automation and AI terminology). These organizations are applying AI across fundamental financial domains rather than relying on individual pilots. We’ve observed that some CFOs and their teams are using AI to more accurately forecast, monitor working capital in real-time, shorten reporting cycles, and uncover new opportunities for cost savings. These initiatives have enabled companies to become more agile, forward-looking, and aligned to the needs of their organizations.
In this article, based on our experience, we examine three areas where finance teams are delivering the most value with AI: strategic planning and control, cash and working capital management, and cost optimization. Each section includes case studies that demonstrate how leading organizations are using Gen AI and agent systems to improve the way their finance functions operate. Finally, we’ll identify five common mistakes that can slow you down and what you need to do to overcome them.
Strategic planning and control: How AI provides better insights
Decision support tools that leverage a combination of predictive analytics and generative AI make it faster and easier to access enterprise data, generate reports, and run predictions and scenarios. These tools support finance leaders and their teams while making data more accessible to decision makers across the enterprise. AI tools typically combine several common features, such as alerts to help finance leaders focus their time and attention, interactive root cause analysis to help users understand factors impacting performance, and alternative scenarios to consider. AI is well-suited for these tasks because it is particularly good at integrating multiple layers of information, such as external, financial, and operational sources, into a consistent view.
For example, at a global consumer goods company, the Gen AI assistant helps finance professionals provide insight into budget variances to business leaders across different departments and markets. This tool replaces manual number crunching, saving financial professionals an estimated 30% of their time.
In another example, a large North American financial institution uses the gen AI tool to generate the first draft of a report documenting internal risk model requirements and updates. The tool also helps generate market-specific risk models by combining internal data and public sources to streamline what was once a time-consuming process.
Across multiple industries, companies are developing and deploying decision support agents enabled by Gen AI and Agent AI to significantly reduce the time needed by finance teams to make resource allocation decisions. Instead of manually pulling reports and piecing together insights across features, these teams now use natural language to generate complex scenarios during planning sessions. These AI tools integrate data from multiple sources, such as customer relationship management systems and financial, operational, and marketing data sets, and surface managerial alerts, such as declining ROI. It also provides root cause analysis (for example, “The problem is caused by cost category A in region Y”). The tool then suggests action steps based on the data (for example, “Based on recent ROI and forecasts, consider shifting 10% of your sales budget to digital marketing to drive higher growth”).
Of course, the specific AI implementation will vary by organization. The few finance departments that have implemented it well have observed a 20-30% reduction in the time finance professionals spend processing data. Use the time you save to act as a business partner to support strategy execution. AI tools also make it easier for finance departments to provide insights across the organization by quickly generating customized reports that maintain appropriate security and hierarchical access controls.
Managing cash and working capital: How AI scrutinizes terms and invoices for accuracy
AI-powered agent workflows enable the next level of automation in both payment and receivable processes, increasing efficiency for procurement and other back-office teams.
For example, a global biotech company implemented invoice-to-contract compliance using an agent AI system that ingested contracts and invoices throughout the year and checked that all contract terms were applied correctly. This approach helps prevent value leakage if vendors overlook or misapply terms such as early payment discounts, tiered pricing, and volume rebates. This runs alongside existing automation, extending coverage to the full scope of a company’s spend base and reducing the need to manually monitor high-value contracts. The system can interpret each vendor contract and its terms, track compliance on incoming invoices, and identify issues that only occur across multiple invoices, such as when cumulative purchase volume qualifies you for a lower price tier.
Using this AI system, the company identified contract breaches that amounted to approximately 4% of total spend (a level of breach that is not uncommon in the industry). This presented a clear opportunity to recover lost value and improve margin performance. To put this in a hypothetical context, for a company with $1 billion in nominal spending, closing that gap could improve recurring income by $40 million.
Cost optimization: How AI analyzes detailed spend to find savings
AI can simplify the time-consuming task of categorizing detailed costs by analyzing complex invoices and purchase orders and organizing them into clear, structured categories. With increased visibility, finance teams can apply advanced algorithms to identify anomalies and areas of waste.
A leading European financial institution set out to identify hidden inefficiencies across its operations to better understand and control its indirect spend base. We started by collecting invoice-level data from thousands of suppliers and organizing it into detailed cost classifications with four levels of detail and approximately 400 subcategories. To efficiently process and classify this data, organizations used a combination of large-scale language models and advanced analytics. With structured datasets in place, cost inefficiencies surfaced by applying both automated and semi-automated techniques (with experts reviewing the output) for anomaly and pattern detection. This analysis revealed specific opportunities to reduce costs and waste in areas such as energy use, travel and transportation, and facilities management. While each category alone resulted in small savings, together they contributed to approximately 10% savings on a multi-billion euro spending base.
Another large European company in the packaging industry is now able to better manage its fragmented supplier base by using Gen AI to categorize over 10,000 suppliers. While management teams have traditionally focused on the top spenders, many smaller suppliers, many of whom are in the indirect spend category, remain poorly understood. The company used gen AI to more accurately categorize all of its suppliers and identify patterns and overlaps that had gone unnoticed before. This enhanced visibility helped uncover cost savings opportunities and optimize procurement strategies. This classification also revealed gaps in supplier diversity, allowing the company to expand its sourcing in underserved regions.
Overcoming barriers to scaling AI in finance
To harness the potential of AI in finance, teams can do more than just add new tools to old ways of working. Core processes, people, and technology must be rewired to ensure adoption sticks and creates value. Along the way, common pitfalls can slow or stall your progress, including:
- Waiting for perfect data. Some teams delay the rewiring process until all datasets are fully accurate, connected, and standardized. In fact, finance teams can create value by providing use cases to process today’s data while strengthening their data foundation.
- Trying to transform all at once. Waiting until the entire feature is “AI-enabled” will slow progress. A better way is to transform domain by domain, building momentum and capacity to deliver sustainable results.
- Jumping in without a clear roadmap. Pilots launched without direction are unlikely to scale. Finance leaders need a roadmap tied to business priorities that allows them to clearly choose which use cases to pursue first and which to work on next. Use cases also need to be supported by technical talent that can contribute to their success.
- Ignoring change management. Often the biggest barrier is not technology but implementation. Strengthening your team and building buy-in is essential to gaining and maintaining influence.
- Automate fragmented processes. Without first simplifying and standardizing core workflows, AI will only increase complexity. You can effectively scale your technology by removing unnecessary steps and making processes consistent across teams.
Avoiding these pitfalls requires a clear vision, strong business alignment, and a focus on actual execution. Finance leaders who approach AI with a strategy rooted in business needs are best positioned to achieve lasting impact.
As AI adoption grows, the difference between short-lived pilots and pilots that create lasting value is becoming clearer. As the case studies in this article demonstrate, successful companies are those that connect AI to specific business needs, streamline core processes, and leverage technology to free up capacity for higher-value work. For CFOs, the message is clear. The opportunity is real, but capturing it requires moving beyond experimentation to disciplined execution based on business priorities.
