Paul Gogarty (Dublin Mid West, Independent) | Oireachtas source

I congratulate Ms Kennedy on her new role and I thank her for the presentation. I was here for it but then I had to go to the Chamber. I will try not to repeat questions that others have asked. I checked with my colleague and not too much has been asked about the AI side of things, so that is where I will focus. I noticed a piece in the Journal of Accountancy, which is US based. It did a survey of attitudes to AI among audit teams. Almost a quarter of the respondents cited a lack of training and infrastructure. Some 17% said that the technology was too expensive or no use and another 13% cited inability to access usable client data. There are a lot of benefits. Mr. Prendergast mentioned that for the larger companies, there should not necessarily be a need to utilise AI, to a certain degree. From my reading I can see that there are benefits. The Journal of Accountancy article states, "... it helps the auditor more appropriately refine the scope and planning of the audit, to focus on procedures that are more responsive to actual risk ..." It can go deeper into datasets. In that context, we have to acknowledge that there are real benefits to AI. Have the witnesses noticed any pitfalls to AI? I would argue that smaller companies might be more likely to try a more basic version, to see if they can get more efficient and that is where the pitfalls might be, because the checks and balances might not be in place. Have the witnesses seen situations where companies have started using AI and while it looks fine and they have tested it a few times, there is actually gibberish underneath? Does the authority issue guidelines to them? When it is conducting its audits, does it ask companies a specific question about whether they use AI and if they do, what have they been using? In that context, does the authority have any educational role aside from the enforcement role?