Mr. Kevin Prendergast:

If the Deputy has an interest in this, and I hope he does, IFIAR, the global body, recently issued a report from its technology task force on the use of technology, including AI. The report very much supports Ms Kennedy's comment that there is a lot of talk, but the level of actual use is probably not really up there yet. The investments are enormous, with billion-dollar sums being invested by the largest firms. It is interesting when we talk to them because the thinking is quite advanced. For example, they do a lot of talking about ethical rules and guidelines on the use of AI, in order to make sure that it does work properly. The Journal of Accountancy is an excellent publication on this issue. The Deputy pulled out one of the nitty-gritty points about useable client data. There could be a wonderful system that says it will look at every single transaction that is sent in and ten seconds later a result will be produced. Almost invariably, one cannot get the datasets to talk to each other and get them all on to one system. There are a lot of real-world, practical challenges to making AI work. That said, the potential upside is very high.

The key point is transparency, so that everybody knows where AI is being used, because the risks change. A junior employee who may not have much experience is one set of risks but AI doing a huge amount of work is a completely different set of risks. As global regulators, the other thing we are looking at is how we change our inspection regime to ensure that we are taking account of those particular risks, and we are very much at the beginning of that.