Artificial intelligence is becoming central to many core banking activities, powering systems used for
A 2025 American Banker
Some institutions may be more dependent on AI than their leaders realize. On paper, continuity plans look robust, detailing manual reviews or alternate processes, fallback queues, rerouting rules, and escalation protocols. Still, are banks sure that, in practice, those paths could be operated at scale with the staffing and skills available today? AI-enabled systems are often used at points in the workflow where large volumes of activity are filtered, scored, flagged, or prioritized before staff see them, and earlier manual or rules-based pathways have been reduced as processes were streamlined around those capabilities.
If an AI system becomes unavailable or its performance degrades, the operational impact can be significant. Fraud detection pipelines can stall, exposing the bank to higher losses, or anti-money-laundering monitoring can miss suspicious activity that would normally be flagged. In credit activities, loan approval processes can freeze, disrupting revenue flows and delaying decisions for customers. In some cases, systems remain technically available, but concerns about output reliability can prompt risk and compliance teams to suspend or restrict their use.
From a continuity perspective, having a clear map of critical models and data flows, including their operational dependencies and any components operated by key third-party providers, can help banks understand how long specific AI systems can remain offline before customer or regulatory impact becomes material. A point worth considering for boards and executives is whether their confidence about operating without a particular AI system is supported by testing and exercises conducted under realistic conditions.
CodeBoxx Academy is filling a void for banks and other companies that desperately need AI experts. Peret's time behind bars uniquely informed how he runs the school, he says.
Supervisors are placing growing emphasis on operational resilience and continuity. In the United States, existing supervisory frameworks already cover AI systems through expectations for model and third-party risk management:
International supervisors are moving in a similar direction. The European Union's
As AI becomes more closely connected to essential banking workflows, continuity thinking becomes an important part of board oversight. This means that processes that depend on AI systems are brought into existing continuity discussions and exercises, so that their operational dependencies and fallback options are understood as clearly as those of other critical systems.
For bank boards, the path forward rests on acknowledging that continuity planning must evolve as AI is integrated into more critical processes. Preparedness is ultimately what will keep the bank running even if AI systems are disrupted.
