Controls for Governable AI in Payments Modernisation
At Sibos, Bridget Hall of ACI Worldwide discussed the progress of instant payments and AI-driven modernisation, following a session talk. The conversation placed the emphasis on controls rather than tools alone, framing governance as the foundation for strong and manageable infrastructure change.
Effective controls begin with clear ownership and accountability. Institutions need defined decision rights for AI-assisted changes, human oversight at critical points, and documentation that allows supervisors and internal auditors to trace how models influence infrastructure choices. Without that clarity, speed can outrun control.
Operational safeguards are equally important. Continuous monitoring, model performance checks, data quality protections, and change-management gates help detect drift or unintended effects. These measures should be proportionate to risk and folded into existing resilience and compliance frameworks, not treated as a separate layer that adds friction without improving outcomes.
Hall's remarks suggest a broader lesson for payments overall. Instant payments raise expectations for speed and reliability, so AI modernisation must be paired with governance that keeps systems explainable and controllable. Strong controls do not slow progress; they make it durable and defensible as adoption expands.
Source: Finextra
