Financial Executives International released the "AI Framework: Internal Control Over Financial Reporting" whitepaper on June 25, 2026, to help companies integrate AI into financial reporting while maintaining effective internal controls. Developed by corporate controllers and chief accounting officers from Fortune 100 and large public companies, the framework offers a principles-based approach to capture AI's efficiency without compromising the reliability that capital markets require.
The document is intended as a living resource that evolves alongside AI technology and regulatory expectations. "As finance leaders race to capture the value of AI, they are being asked to do so without disrupting the reliability of financial reporting that markets depend on. This framework reflects exactly what FEI does best: convening the most respected practitioners in the profession to translate an emerging challenge into practical guidance. It gives finance teams a credible, principles-based starting point to adopt AI responsibly while protecting the integrity of their internal controls," said Andrej Suskavcevic, President and CEO of FEI.
Rather than auditing the inner workings of black-box AI systems, the framework shifts focus to human oversight and outcome-based validation. It outlines four control approaches that companies can deploy individually or in combination.
Four business control approaches
- Human-in-the-Loop (HITL): Human intervention at strategic decision points to validate outputs and manage model uncertainty.
- Performance Testing: Curated test data sets compare AI outputs against known ground truth to verify accuracy.
- Multi-Model Validation: A primary AI system is compared against an independent challenger model to identify variance.
- Data Analytics: Advanced monitoring techniques analyze trends and detect anomalies, ensuring the AI operates within established limits.
The framework also addresses AI-specific risks such as hallucination, model drift, and algorithmic bias, and provides guidance on SOX scoping and risk assessment consistent with the COSO Internal Control - Integrated Framework. It includes an outlier management and resolution process to handle unexpected results.
"AI doesn't change management's fundamental responsibility for effective internal controls - but it does change how we earn comfort over them. By focusing on validating outcomes, the framework gives companies practical control approaches they can put to work today, and a path to scale automation responsibly as their AI systems prove reliable over time," said Aaron Anderson, contributing author and CCR member.
The whitepaper is not legal guidance. Readers should consult legal counsel for decisions that may raise liability concerns. The complete Framework is available for $195 on the FEI website.
Why this matters for finance leaders
With AI moving from experimentation into core accounting processes, controllers and chief accounting officers need concrete methods to manage risk without slowing innovation. The framework delivers actionable steps for scoping AI systems, validating outputs, and maintaining SOX compliance. Finance professionals who want to build practical AI skills can follow an AI Learning Path for Accountants that covers automation and financial reporting tools. As adoption accelerates, targeted training in AI for Finance helps teams stay ahead of the curve.
Your membership also unlocks:
