SBS Chief Says Data, Not Models, Is Stalling Bank AI Pilots

Banks are finding that the biggest obstacle to deploying artificial intelligence is increasingly the quality and governance of their data, rather than access to capable models. An SBS executive said fragmented records, inconsistent data standards and unclear access controls can prevent financial institutions from moving AI projects beyond the pilot stage, highlighting the importance of data engineering, compliance and governance.

SBS recently said data preparation and governance, not model performance, are the main reasons bank AI pilots stall. The report did not identify individual banks or disclose project budgets or an exact announcement date. The assessment suggests lenders seeking to scale proofs of concept into production must first improve data cleaning, integration, permissions and quality monitoring.