Has anyone in Audit seen AI make something better?
retained text
Been working in Audit at a Big 4 for a couple of years and always considered the tech savvy person at the office. Recently, there has been a huge push by managers and partners to implement AI. Phrases like "how can we use AI to make this better/faster/more efficient" are thrown around almost daily. Of course, this comes as clients request us to use AI for a more efficient (cheaper) delivery. Having sort of a technical background, I've tried the various LLMs (including our in-house model) a fair bit without any tangible results. Aside from the DataSnipper-like applications, I struggle to see the value in "AI". Sure they can "summarize documents", "improve e-mails" etc but in my experience this process takes longer than just writing on your own with little to no improvement from AI typing. The only use case I myself have found (as well as many of you) is to use LLMs to help write Macros, VBA, miscellaneous scripts. However, I find this kind of useless in audit as you tend to work with multiple clients, multiple work books, which eliminates efficiency gains from script-like automation. Further, most people who studied business/accounting/finance barely know how to write an XLOOKUP, so I doubt that they are able to even conceptualize and structure a problem for an automation script. Obviously this is a polarizing topic, and I am not interested in stirring up an argument. I am just interested in your perspectives. Have you found any use in AI? Is this just a case of managers selling something that they don't understand? TLDR; Can't find a valuable use case for AI in audit.
document fetched 2026-09-06 00:17 · canonical https://reddit.com/r/Accounting/comments/1w83bon/has_anyone_in_audit_seen_ai_make_something_better