Autonomous finance demands accuracy, explainability, and accountability
Expert argues AI in finance requires higher standards for accuracy and auditability, not lower—challenging the notion that automation simplifies accounting work.
TLDW: AI could fundamentally reshape audit firm economics by enabling one-person practices, but firms must first document workflows before automation to capture real productivity gains. Key points: - Blake and David discuss whether AI could make single-person audit firms economically viable, shifting the traditional partnership model - AI-powered continuous audits are emerging as a practical application that can reduce audit cycle time and improve real-time financial monitoring - Firms need to document existing workflows and processes before attempting to automate them, otherwise automation efforts will fail or miss optimization opportunities - The podcast covers Mercury's new general ledger product launch as a signal of evolving accounting infrastructure for the AI era - Tax evasion case study (OnlyFans creator with $3M unreported income) highlights ongoing compliance challenges that AI detection systems could help address
Continue reading
Attention: This is a machine-generated transcript. As such, there may be spelling, grammar, and accuracy errors throughout. Thank you for your understanding! David Leary: [00:00:04] A 39 year old Connecticut OnlyFans creator, just pled guilty to tax evasion after earning over $3 million between 2019 and 2022 without paying a dime to the IRS. See more from Stanford. Didn't just forget to file. She actively tried to hide her income by opening 19 different bank accounts and shuffling money between them. Coming to you weekly from the OnPay Recording Studio. Blake Oliver: [00:00:35] Hey everyone, and welcome back to the Accounting Podcast, your weekly roundup of news in the profession. I'm Blake Oliver. David Leary: [00:00:41] I'm David Leary. Blake Oliver: [00:00:43] David. Halloween's coming up just a little over a month away. You got your costume picked out yet? David Leary: [00:00:49] Yeah. I was thinking about being the only accountant who lost their job due to AI. I'm like, I'll dress up as, like, a unemployed accountant and blame AI for it. That'll be the only one. Blake Oliver: [00:01:01] Halloween spending is going to reach 13.5 billion this year. It's like $115 a person. So that's your budget. That's your budget for your Halloween costume, plus your candy. I wanted to. I'm trying to come up with some like good Halloween costume ideas. So I ran them through a clod and it connected to Canva and gave me some. And I was actually pretty impressed at what it came up with. Let me show you now and our listeners here as well. Um, can you guess what this one is? This is the ghost of the 150. David Leary: [00:01:35] That's so awesome. That is so good. Like that is outstanding. Like that's so great. Blake Oliver: [00:01:41] This is an easy one, right? So, uh, it's just a, uh, you know, white sheet, obviously with the holes cut out for your eyes and it says little, you know, I actually don't know what is what is on here. David Leary: [00:01:53] It's like a tombstone. Blake Oliver: [00:01:54] Um, well, there's, there's a sign that says the The 150 credit hour. And then the ghost is holding a Master of Accountancy diploma. And, uh, yeah, so there's that one. Um, we've also got, uh, this one here. This is the private equity, uh, CPA firm flipper. So, you know, you dress up in a construction outfit with the hard hat and the flannel shirt and some, some tools. Uh, interestingly, in this photo, I just noticed it. Now he's holding a, like a one of those construction hammers, but also a, uh, a spatula as a flip flipping. David Leary: [00:02:34] He's flipping. Blake Oliver: [00:02:34] Yeah. And he's got a sign on that says, uh, CPA firm sold, sold, sold now 14 times EBITDA. That's great. There's that one. Okay. And here's another one. Uh, this is the IRS skeleton crew. So skeleton outfit with an IRS, uh, I don't know, like name tag. David Leary: [00:02:57] Name, badge. Blake Oliver: [00:02:58] Name, badge, and holding like an inbox, like one of those metal wire inboxes that says almost done with a stack of papers in it. And then this is actually a great, uh, double costume, right? A couple's costume. Uh, the other costume is the revenue agent. It's a robot made out of cardboard boxes. It says revenue agent beta, and it's got like the, you know, um, old school, like 1950s, kind of like robot thing going on with the, I don't know how you describe it. Like the remember the Cylons from the original Battlestar Galactica? David Leary: [00:03:32] It's it's, it's dryer tubing for the arms. Yeah, yeah. Blake Oliver: [00:03:36] Yeah, yeah. So those are some, um, Halloween costume ideas for our listeners. Let me know if you come up with any good ones. I am looking for something to do at Intuit Connect coming up, which is always the week before Halloween. So it's there's always a costume party. It's fun to dress up. And if you're going to be at Intuit Connect, let us know as well. We'd love to meet up with you. Um, we will be there. All right. David. We need to get into the news this week. You've got a story about, like, the AI bubble. David Leary: [00:04:08] I got a scary story. Yeah. Blake Oliver: [00:04:10] And you've got some tips for rolling out AI in your firm. Practical tips are always helpful. Mercury has launched a GL and an OnlyFans model who didn't pay $1.1 million in taxes. That might be the largest I've ever seen in terms of a tax bill that somebody didn't pay. Like, just like a, an individual, right? Not a company. David Leary: [00:04:34] Yeah. And the owner, like a billionaire, like this is not a billionaire. Blake Oliver: [00:04:37] Yeah. This is just yeah. Um, well, she did make, what, 3 million, over $3 million. Um, I've got a story about a new EBITDA adjusted EBITDA metric. Uh, those are always fun. And Christina Ho, the Pcob board member who we have interviewed on the show about issues with audit and the Pcob. She's got a new A AI audit startup and just signed her first audit report. That's cool. And I don't know if we're going to get into any more of it, so I won't tease stuff that we can't get to. So before all of that, David, let's thank our first sponsor. David Leary: [00:05:13] Yeah. So our sponsors today, we have digits Thomson Reuters on pay and cloud accountant staffing. Let's be honest, accounting software hasn't changed much in decades. The prices keep going up, but the software still expects you to do all the work. Digits is different. Digits is the world's first AI native general ledger with built in AI agents trained on your firm's standards across every client. In your book, they code transactions, prepare schedules, reconcile accounts, run quality checks, even chase clients for open items. So your team moves out of prep into reviews, advisory and the work that actually grows the firm. And because everything runs inside one platform, your ledger, reconciliation schedules, reporting bill pay, client collaboration. There's no more switching six tools to stitching, six tools together just to close one client's books. Firms on digits are reporting a 70% gain in workflow efficiency, shipping annual cleanups in days instead of weeks, and running monthly bookkeeping in 1 to 3 hours per client accounting software that actually works for your firm. To see why hundreds of firms are making the switch to digits, head over to The Accounting Podcast dot com slash digits. That is accounting podcast.com/digits. Blake Oliver: [00:06:25] And please do use those links when you visit our sponsors websites. It helps them know that you found them by listening to the podcast, and we really appreciate that. All right, David, let's talk about the AI bubble two big to fail or no. Two interconnected to. David Leary: [00:06:39] Interconnect to fail. And I saw that headline in my brain instantly went to too big to fail. And what was that from? Blake Oliver: [00:06:46] That's from, uh, the Great Recession. The who could forget the mortgage crisis? David Leary: [00:06:50] 27 2028. Right. And we've kind of talked about this before. We've made this comparison to the AI bubble and the interconnectedness of all these AI companies. Well, Sona Asset Management has just done a released a report and there's a slide we can share. It's it's a little dim, but I'm going to still share it anyways. And it really shows the circular. The circular. I want to say circularity, but I don't know if that's the correct word. And so you have these companies. You have the hyperscalers, the model developers, chip suppliers, the NIO clouds. And they're all selling to each other and to kind of. So what they did in this report, there's 255 public companies in the AI ecosystem now worth $50 trillion, combined with a $6 trillion in debt. And, you know, a lot of it is, uh, the revenues concentrated. So core we've got 67% of its revenue from Microsoft applied digital gets 56% of its revenue from Oracle and Core. We've. And so you start seeing this interconnectedness and we've been stacking this and talking about this over the last couple of weeks. Right. Microsoft has $80 billion in accounts receivable. So now the real numbers are coming out. But when I saw the headline, it really set in about 2007, 2008, the mortgage crisis. So I kind of went back to get a feel on the scale. So there were about 1 trillion subprime mortgages, mortgage mortgages. Oh my goodness. One more try. 1 trillion in subprime mortgages of the 12 trillion US mortgages at the time. So 1 trillion of 12 trillion. So now you have AI company debt is 6 trillion, and half of that debt is circular, 3.6 trillion is interconnected. Blake Oliver: [00:08:38] 3.6 over half of all that debt. So it's about what half. You said it was 12 trillion in mortgage. David Leary: [00:08:46] Yeah. So so that was about just under 10% was in subprime mortgages. Blake Oliver: [00:08:52] Okay. So one you said one out of 12 in this. Okay. Just under 10%. And in this case, we now have it's like half the total value. But more than half of that is interconnected. David Leary: [00:09:05] Interconnected. Blake Oliver: [00:09:07] Oh, wow. This could be worse than I thought. David Leary: [00:09:08] Yeah. So if you think about how a trillion took out Lehman just a trillion and that required government intervention, caused recession. Right. And I see this article and it makes me start thinking like, are we being primed and trained for a government bailout here? And like, let's step back and let's just let's pretend they don't all collapse. We don't have the worst case scenario, right? Let's say all they do is pull back on data center construction. So they pull back on data center construction, or maybe worse than that, they don't pay construction companies or contractors. Blake Oliver: [00:09:43] Okay. David Leary: [00:09:45] This is going to quickly spread across construction firms, Electricians, Hvac, q Hvac equipment suppliers, truckers. Right. They're going to be stuck with unpaid bills. Construction companies. And that's essential