Evolving finance for the AI era: Key issues for controllers

Controllers are evolving beyond their traditional roles of compliance and control, leveraging advanced analytics to influence business decisions and shape enterprise strategy. As AI redefines what’s possible, investment in the right technology, data and talent is essential.

A particularly promising frontier is agentic AI — autonomous, intelligent systems designed to perform complex tasks. For controllers, these systems may offer a powerful opportunity to deepen their impact across the organization.

However, integrating AI into finance is easier said than done. Controllers continue to navigate daily demands while determining how to realize the full potential of AI. In a recent PwC Controller Agenda webcast poll, 60% of the more than 10,000 finance professionals who registered said they are not yet using AI in finance. Among those who are, 22% said they were using AI for financial reporting and closing, while 20% said they used it for automating accounts payable/receivable.

So where should controllers start? Here are five key issues that should be on every controller’s agenda now.

AI is changing the nature of work itself. Amid ongoing talent shortages, AI offers an opportunity to free up staff from repetitive tasks, allowing them to focus on higher-value, strategic work. To remain competitive, it’s critical to reimagine roles, responsibilities, training and recruiting.

Help your people to master AI tools effectively and become more adept in reviewing outputs. Enhance training around data governance and Responsible AI practices. Beyond new skills, encourage workers to embrace AI and innovate with it. This can require a cultural shift, with leadership providing incentives and making it clear to staff that people with these skill sets will always be important.

Anthony Abbatiello, PwC Principal and Workforce Transformation Leader, shares how controllers can lead strategically by rethinking work, reshaping teams, and empowering people for the AI era.

AI agents are helping to transform how data is gathered and analyzed. These software agents can be designed to execute multi-step processes within parameters defined by humans. AI agents — whether embedded within enterprise platforms or built from software development kits for specialized purposes — are capable of handling tasks, suggesting outcomes and orchestrating entire workflows.

Consider revenue recognition. An AI agent could ingest accounting literature and company policy documents, interpret it in the context of organization-specific scenarios, and offer recommendations aligned with GAAP requirements. Importantly, this still requires human oversight to make sure the agent is getting it right — and if not, to further refine the model so it improves.

AI highlights many core issues controllers deal with every day — including data quality. According to PwC’s 2025 Global Digital Trust Insights Survey, 48% of business executives reported that they’re prioritizing data protection and data trust investments over the next year, ahead of technology modernization and enhancement. Without a strong data foundation, tools and processes likely will be inefficient. While organizations have long grappled with data quality, the expanded use of diverse and complex data sets in AI-enabled processes underscores the importance of enhanced governance and controls to help manage evolving risks.

However, implementing AI doesn’t require data perfection. Agentic AI is more adaptable than traditional automation. It’s capable of flagging anomalies and working across disparate systems. Whether you use a single ERP with edge systems or multiple reporting tools, AI agents can pull data from various sources through a cloud-based stack and deliver analysis that supports decision-making.

Start by identifying outcomes. Then, collaborate across finance, IT and vendor teams to build an AI system that can achieve your goals and provide a clear audit trail.

For controllers, establishing explainable, repeatable processes is second nature — and those same principles apply to AI. Responsible AI requires a proactive approach to risk assessment, control development and compliance.

While it may be tempting to wait for regulatory clarity around AI, finance leaders should take the initiative now. Controllers are uniquely qualified to drive governance efforts that will shape effective, trustworthy AI implementations. Partner with IT to establish policies and procedures that uphold data integrity, accountability and transparency.

Return on investment is an integral part of determining when and how to pursue modernization, and yet it can be difficult to demonstrate ROI when implementing technology. For years, finance transformation has been driven by cost reduction, but we’re starting to see a shift on that front. In a benchmarking study, PwC recently looked at the top quartile of companies and found that finance costs as a percentage of revenue have dropped from about 1% over the past 15 years to .76% — and have been flat for the past three years.

For the first time, many companies are spending more time on insights than transactions, and yet controllers still feel the pressure of closing the books. The key is to use technology to accelerate insights without compromising accuracy.

Controllers should strive to articulate ROI in broader terms — showing how finance transformation enhances reporting, improves decision-making and contributes to business strategy. It’s not just about reducing cost; it’s about increasing impact.

Agentic AI can be a powerful tool as the controller’s role continues to evolve. By focusing on embedding Responsible AI practices, strengthening data foundations, reimagining how work gets done and expanding the definition of ROI, controllers can confidently deliver new insights faster. This will require careful planning, collaboration across domains and finance expertise — all of which the controller is uniquely suited to provide as an ever more strategic business advisor.