AI in Finance should not be treated as an IT project
In this interview by Angelika Breinich-Schilly, Christoph von Klimesch, Partner at Eight Advisory, explains how finance departments can truly scale AI by focusing on clear objectives, high-quality data, and robust governance.
Von Klimesch notes that many finance teams approach AI from a technology-first perspective, piloting projects without a defined business problem. He stresses that success comes from identifying the process to improve and setting measurable goals before selecting technology. Investing in data quality and integration is essential, as pilots often fail when moved from clean test data to fragmented real-world systems.
He emphasises that AI should be seen as a finance-led transformation, not an IT project, with realistic milestones and the willingness to stop pilots if targets are not met. Reliable data architecture and full traceability of data sources (data lineage) are crucial for trustworthy results and regulatory compliance.
Effective AI governance means ongoing monitoring, automated alerts for performance issues, and clear escalation paths. Finance staff do not need to become data scientists but must be able to critically assess AI outputs and ensure results make economic sense.
A strong security framework is needed to protect training data and prevent feedback loops where models reinforce their own errors. Cultural adoption requires training, incentives, and clear communication about changing roles, addressing concerns about job security and control.
Read the full article by Angelika Breinich-Schilly, published in Springer Professional on 23 March 2026.
