A third of banks do not maintain logs for GenAI models

A recent Risk Benchmarking study reveals that a third of banks fail to keep logs for their generative AI (GenAI) models, raising significant concerns about oversight in a sector where data integrity is crucial. Larger banks tend to focus on high-risk applications, but this selective scrutiny could leave many gaps in monitoring and accountability. This lack of systematic log review is alarming given the potential for GenAI to inadvertently perpetuate biases or introduce vulnerabilities, underscoring the need for more stringent regulatory oversight and best practices in AI governance.