UK Finance has published a blog looking at the implications of frontier AI for financial stability. It notes that the BoE’s latest Financial Stability Report highlights the potential for frontier AI to change the speed, scale and economics of cyber risk. The latest models can carry out multi-stage attacks with little human input, but what is less clear is whether they are learning to compromise well defended targets.
So defence is still key, and faster vulnerability cycles may lead to firms needing to patch more often, which will bring both cost, testing constraints and potential outages and change fatigue. While the CTP regime is important, the risk doesn’t stop with these providers – the resilience of all third party providers is key to financial institutions’ defences.
The blog concludes that the risks aren’t new, but the speed and scale of them is, and this is what firms will need to manage.