Supervisory findings
The Basel Committee on Banking Supervision published an October 19 newsletter describing data challenges in assessing money-laundering risks and the importance of human oversight when banks use artificial intelligence to monitor transactions. [1]
A stocktake across 19 jurisdictions found that supervisors use questionnaires, financial-intelligence reports and audits to assess risks and controls. Manual processing remains common, while incomplete or poor-quality data complicates the work. [1]
The newsletter also reports discussions from a November 2024 industry workshop. Participants described potential uses for AI in spotting unusual activity, grouping customers and adjusting monitoring thresholds. They stressed human involvement in validating models, reviewing alerts and making final decisions about suspicious transaction reports. [1]
Participants identified data quality, algorithmic bias and resource demands as obstacles. The publication shares supervisory observations and workshop views; it does not announce a new binding rule or establish that AI systems prevent financial crime. [1]