Finance Leaders Shift From Piloting AI to Full-Scale Deployment
Ninety-three percent of US companies will deploy or scale AI in their finance functions within the next 18 months, according to a KPMG survey released today. The shift marks a fundamental change: finance leaders have moved past pilot projects and are now building multi-agent AI systems across entire workflows.
Half of surveyed companies are already planning to orchestrate or develop multi-agent systems. Two years ago, the focus was testing. Now it's orchestration.
"This moves the goalposts for our profession," said Christian Peo, KPMG US Vice Chair - Audit and Assurance. "To maintain trust in the capital markets, the auditor of the future will have to both audit financial statements and provide assurance over the AI systems that help produce them."
The report surveyed 1,013 senior finance leaders across 20 countries and 13 sectors, including 163 US finance leaders.
ROI is meeting expectations-but people remain the barrier
Nearly three-quarters of companies report that AI initiatives are paying off, with 46% meeting ROI expectations and 28% exceeding them. Among those disappointed, the top barrier is slow organizational adoption and change management.
This reveals a core problem: AI success depends as much on managing people as managing technology. Finance leaders cite two main obstacles to building AI literacy across their teams. Sixty-four percent lack clear, role-specific use cases. Sixty-one percent lack hands-on practice environments.
The implication is direct. Practical, targeted training is essential to AI transformation. Without it, even sophisticated systems underperform.
The next wave: predictive insights and natural language coding
Finance functions see the greatest opportunity in generating faster, predictive insights for decision-making, cited by 45% of leaders. This push for sophistication extends to emerging techniques.
Twenty-six percent of companies are evaluating "vibe coding"-using AI to convert natural language into working code. Nearly half are already piloting or actively using it.
"We're seeing AI democratize technology development, allowing domain experts to build their own solutions," said Brad Brown, KPMG US and Global Chief Digital Officer, Tax. In a recent pilot, tax professionals with limited technical backgrounds coded working software prototypes in weeks.
Assurance becomes mission-critical as AI risks grow
As companies embed AI, their concerns are shifting. Fifty percent worry about cyber and AI-native security threats. Forty-eight percent focus on the accuracy of AI-generated financial outputs.
This has made independent assurance a prerequisite for innovation, not just a defensive tactic. Ninety-four percent of organizations already rely on third-party assurance providers.
Finance leaders find this specialized assurance most valuable for three areas: data security and privacy (60%), ensuring AI model performance and reliability (56%), and navigating regulatory and compliance requirements (53%).
"A cyber threat to an AI system is now a direct threat to the accuracy and completeness of the financial information it produces," said Matt Johnson, KPMG US AI Audit and Assurance Leader. "Independent assurance is therefore increasingly critical and complex."
For finance professionals looking to build AI competency, resources like AI for Finance and the AI Learning Path for CFOs offer structured approaches to understanding both deployment and governance of these systems.
Your membership also unlocks:
