Latest / The Accounting Podcast / Fake SOC 2 Reports as a Service & Games for Busy Season

Transcript

Fake SOC 2 Reports as a Service & Games for Busy Season · The Accounting Podcast

• There may be errors in spelling, grammar, and accuracy in this machine-generated transcript. David Leary: [00:00:04] Is it like working at an accounting firm during the busy season? Just the escape room already. Like you have to solve all these puzzles or you don't get to go home, like, right? Like, yeah, right. What's the difference between an escape room and busy season at a firm coming to you weekly from the OnPay Recording Studio? Blake Oliver: [00:00:25] Hello and welcome to the accounting podcast, your weekly roundup of news in the profession. I'm Blake Oliver. David Leary: [00:00:31] I'm David Leary. Blake, you got me with your headline today. I'm so interested in this fake Soc2 report story. Blake Oliver: [00:00:38] Oh, this is wild. Thank you to Jacob Schroeder for sending this my way. There's this compliance startup called delve that does Soc2 reports, and there's allegations that they have sold hundreds of companies, fake Soc2 reports, certifications. This is something CPA firms do by generating auditor conclusions, fabricating board meeting notes and using Indian certification Mills operating through empty US shell companies. It is wild and we are going to dig into that. But first I want to talk about what happened this morning at LaGuardia airport, because we've got basically follow up on the story we've been covering about FAA budget cuts, and it ties into this whole TSA thing and the federal government budget impasse and shutdown. But first, David, let's thank our sponsors who are the sponsors for this episode? David Leary: [00:01:32] Sponsors. This week, we have Digits Cloud Accountants, staffing on Pei and UNC Kenan-flagler Business School. Blake Oliver: [00:01:39] Let's thank digits first. Let's be honest, accounting software hasn't changed much in decades except for rising costs and declining service. Now there's finally a reason to switch and never look back. You can now do your bookkeeping on digits. Digits is AI native accounting software that works for you, not the other way around, while other platforms just slap ChatGPT on old workflows and call it AI. Bookkeeping demands more than a chatbot. It demands precision, auditability, and trust. Digits rebuilt ledger software from the ground up with probabilistic categorization and human review prompts the result. Over 95% of transactions auto book with unmatched accuracy, 54% better than ChatGPT style models so you can close faster, stay in control, and finally, stop wrestling with your accounting software. Low value clients are now high value for years worth of cleanups. Now take just two hours, all while your clients get visually stunning reports, streamlined collaboration and insights they need to make better decisions. To see why hundreds of firms are making the switch to digits, head over to The Accounting Podcast. That's The Accounting Podcast ADP Accountant Connect d I g I t s. And now, David, let's talk about this breaking news. Welcome to our livestream viewers. We are recording now on Monday. And this morning there was a crash at LaGuardia airport. Why are we talking about this on the show? Because it's something that we've talked about in the past. It's f a, a staffing, which is directly related to the federal government and the operation of it, which includes, uh, budget cuts and reporting from The New York Times says that even though this is a breaking story and we're just getting information about this air traffic audio appears to show that the LaGuardia tower was dealing with an incident before the crash. And a controller, an air traffic controller appeared to be distracted when the Air Canada jet struck a Port Authority fire truck, killing two pilots. Uh. A flight attendant was ejected from the plane. Dozens of people were taken to the hospital, many seriously injured. What happened? David Leary: [00:03:58] Several minutes on the ground. So obviously fire trucks on the ground, the planes taxiing either just landed or just about to take off. Blake Oliver: [00:04:05] Yeah, it was, uh, I can't I don't know if it was landing or taking off, but there was traffic on the runway and the plane collided with the fire truck. And several minutes after the collision. This is according to the New York Times. A controller at the airport told the pilot of a frontier jet that the airport was closed until further notice. And that, quote, we were dealing with an emergency earlier, unquote. It's unclear what emergency the controller was referring to. The executive director of the Port Authority, Kathryn Garcia, said that air traffic controllers were responding to an aircraft whose pilot, after multiple attempts at takeoff, had requested help before the Air Canada crash. And it's not clear how many controllers were in the tower. This was at 11:37 p.m. local time on Sunday night when the crash occurred on the audio stream, there was confusion. The controller was yelling stop! Truck one stop to the fire truck just seconds after giving it clearance to cross runway four, where the Air Canada plane had just landed. So it had just landed and it was slowing down, but it couldn't stop the FAA. The Federal Aviation Administration is trying to determine how many controllers were in the tower on Sunday night. That will be part of the NTSB investigation. According to The Times, aviation experts said the tower likely had fewer controllers on duty than they would have had hours earlier, when the airport had more flight traffic scheduled. And I brought this up today because we've talked about this in the past, the when it comes to the federal budget, the FAA's own workforce plan says that controller staffing had been harmed by repeated disruptions, including the 35 day 2019 shutdown, and that interruptions to hiring and training can have, quote, significant long term impacts, unquote, on staffing levels. Blake Oliver: [00:06:00] The Government Accountability Office says that the number of or the Gao is that accountability or accounting. I don't I don't remember they said that the number of US controllers has fallen about 6% over the last decade, even as flights rose 10%, leaving shortages at critical facilities. We talked about this when it was in the news about the air traffic controllers not getting paid during government shutdowns. It's had a it's been a problem. And I, I, I have been thinking as somebody who flies a lot and I know many in our audience travel around a lot. Hey, you know, if you're an auditor, you're going all over the country. I'm worried about this. And I thought, when is this going to have a real world impact? When are we going to have when are lives going to be lost? And there's not any evidence directly now that this is tied to the the shortage. But it seems to be looking that way from the early reporting. And it's just it's a, it's proof. It's it's proof that or it's a story that shows how the numbers have real world impact. And I would hope that like Washington could get its act together because the impasse, you know, all this stuff, it's not just politics. It has lives at stake, especially when it comes to federal aviation and aviation in general. David Leary: [00:07:26] This time, they're not paying the TSA people, so they're calling in sick, not going to work. But if I remember correctly, years ago when you covered the story, the air traffic controllers were just not going to work because they weren't being paid or they were being forced to work for free. Maybe that was it, I think. Blake Oliver: [00:07:40] Yeah, they, they're required to work by law, but they're not getting paid. And the same thing happens with TSA, right? But what happens is people call in sick, right? They protest this like, who's going to go to work and not get paid? David Leary: [00:07:52] Exactly right. Whatever kind of effort are you going to put in? Yeah. Blake Oliver: [00:07:55] I mean, it's just it's just outrageous. Welcome to our livestream viewers. We've got boring accountant. Hey, how are you doing for coffee emojis there. Welcome, Oswego shark says morning, guys. Terrible news about the airplane crash. Yes, I agree, Tate made the live stream. Great to see you Tate. And we've got Sharif here as well. Great to see you. Um, please put your comments in the chat. Let us know what stories you are following, what we should be paying attention to, what you think about what we're talking about today. And don't forget, you can earn free continuing professional education for watching this episode and listening to episodes of the Accounting Podcast. Get the free earmark app, go to earmark app in your web browser, or get the free app from the Apple App Store or Google Play Store. It's free to create an account and free to earn one CPE Every week. All right. Let's talk about this compliance startup the SoC two reports delve. This is wild. Thanks, Jacob Schroeder for sending this my way. And I want to before I dig into it, just say that these are allegations. I have not independently verified any of this. This is a Substack report very in depth by an anonymous poster. So we should take this with a grain of salt. David Leary: [00:09:14] And so to rewind, I remember, you know, three, four years ago we started seeing these companies. If I have a startup, I need to be SoC two compliant. I could either go find some accounting firm, do this for me. But then these disruptive startups started to show up on market. And I could, you know, for a third of the cost or a 10th of the cost, probably I could use the SaaS product and get my SoC two without ever having to actually hire an accounting firm and paying extra for this. Blake Oliver: [00:09:41] And theoretically, they're still the same thing, right? You're just wrapping these, these startups are wrapping the product that a CPA firm, a traditional CPA firm, would provide into something that's more like a product that you can buy and that you can do online and you can do quickly, sometimes in as little as two weeks. Some of these startups are advertising and there's a lot of them. David Leary: [00:10:04] Oh, it's, you know, budget friendly SoC two. There's like, there's a lot of these companies I know popped up over the last few years. Blake Oliver: [00:10:10] There's a bunch of them and one of them. David Leary: [00:10:11] Is, yeah. Blake Oliver: [00:10:13] One of them's called delve d e l v e. And the allegations in this article by an author called deep Delver say that delve, which is a VC backed compliance automation startup that does C SoC two reports, is creating the appearance of compliance without the underlying substance. According to the author, delve will pre-populate policies, evidence and even auditor facing conclusions and then route customers through audit firms that allegedly rubber stamp reports. The article argues that this leaves customers falsely believing they are compliant with frameworks such as Soc2, ISO 27 001, HIPAA and GDPR while actually exposing them to regulatory, contractual and reputational risk. The article says that delves customers include startups and larger companies handling sensitive data, including P h I, which is health information and personal data. And if these allegations are true, then the affected companies, they could be misrepresenting their security posture to customers, investors and the public. So let's dive into the allegations here. The author says that delves platform is creating fake and misleading compliance evidence. It's built around pre-populated forms. Default policies templated risk assessments, pre-written board minutes, canned security simulations, and manual screenshot uploads presented as integrations. And according to this article, customers can move quickly just by adopting defaults. Even when the defaults describe controls or activities that never occurred. Examples include board meetings, board meetings that never happened, security simulations that were not performed, trust pages showing controls as implemented before any work was done, and policies claiming capabilities like MDM, mobile device management, vulnerability remediation, backup testing, VPNs, virtual private networks, or intrusion detection that the customer doesn't actually have. David Leary: [00:12:35] So, so instead of just getting a checklist and be like, okay, now I have to go execute all these things, which I've done. I've been on that inside as an employee, and you have to do this extra compliance stuff because there's an audit happening for SoC two or something. And so now they just as a service, they just provided all these check boxes done for you. Blake Oliver: [00:12:52] The templates that represent what you really should have done in real life, but you can go through without having to actually do it. David Leary: [00:13:01] It's just done. Yeah. Blake Oliver: [00:13:02] Nobody's verifying that this was actually happened according to the the author. So they're also alleging that delve is pre generating report sections that should reflect independent auditor judgment. Remember these reports, the SoC, two ones in particular. These are AICPA attestation engagements. So they're saying that delve is pre generating auditor conclusions test procedures test results and final report language. The author got Ahold says they got Ahold of draft SoC two reports that show near identical wording across hundreds of client reports. Auditor conclusions they're already embedded before clients filled in. Company specific sections, and this is shown through repeated grammar quirks, identical control language across reports. The same test conclusions appearing in every report, including claims that are seemingly impossible, such as no incidents or no customer terminations and no major personnel changes across clients. And the author says that this pattern, it indicates centralized template generation rather than individualized audit work. Because if you've done the individual audit work, they wouldn't all be the same like this. The author claims that delve hosted trust pages were fully populated and publishable before customers completed compliance work, and that these pages listed controls as completed or monitored live even when they were not implemented. The article has an appendix that has that claims analysis of 322 public Delve Trust pages, with 321 of them all, but one of them showing the same soc2 control set, which seems to undercut Dell's marketing claim that the programs are customized for each customer. So delve is marketing itself as this AI native compliance platform. But in practice, what it looks like it's is that it's static templates, manual forms, screenshots, and a generic chat bot. David Leary: [00:15:23] It's a fake it till you make it again, like you see all the time with a lot of these startups that come in. Blake Oliver: [00:15:29] The author also says there's some sort of like publicly accessible Google spreadsheet with hundreds of draft audit reports and confidential client information. And there are links in this Substack report of an archived copy of the spreadsheet, a large repository of the leaked reports and article backups. So if you want to dive into this after we're done talking about it, I'll put the link into the, uh, into the chat and you can go take a look. So this doesn't really surprise me because when I first heard about these systems, when I first heard about these startups doing this, I thought it would be really easy to fake this kind of stuff, especially if you like use AI to generate what should be in the report. And then how do you actually get firms that are going to sign off on this? Right? That's the hard part. Well, if you think about it right, it wouldn't be that difficult in order to like get a firm to sign off, you just have to get a CPA firm Registered to do these reports to like just sign off, right? So all you have to do is find a firm that's willing to do it without actually doing the work. And we've seen this before. Firms doing actual like financial statement audits. David Leary: [00:16:49] Right? Borgir's. Blake Oliver: [00:16:50] Yeah. Bf Borgir's right in in Colorado. Was it who was doing these for years? I mean, this is the problem with assurance. If you, if you have like a few bad actors willing to just sign off, sign off, sign off, they can make a lot of money. And when are they ever going to get caught? And if they get caught, what happens? So it's very easy to take advantage of the system that we've created as CPAs to game the system. You find a few bad actors, you get them to sign off on all the reports. So I am, uh, I'm, I'm really interested in seeing what happens here. I hope that the authorities will investigate, but I'm also wondering who's going to do this. Is the AICPA going to look into this? Would it be like a state board of accountancy that does this? Who can actually figure this out? David Leary: [00:17:42] Well, just because I'm definitely not an expert on Soc2 in any way, shape or form, but my understanding is the logo you get is an AICPA logo, right? Blake Oliver: [00:17:51] Yes. David Leary: [00:17:52] Like, like you're, you're, you're getting a stamp of approval from the AICPA. So in theory, any companies providing this service are getting some sort of blessing from the AICPA. Either they at least have to register to use the logo. You just can't declare somebody a Soc2 compliant. Just give them that badge. The AICPA obviously is not going to just let that happen without you paying the AICPA, right? Some sort of developer fee or whatever you want to call it. Blake Oliver: [00:18:18] So, so to your point, David, these Soc2 audits, they have to be conducted by an independent, licensed, certified public accountant firm, CPA firm accredited by the AICPA. So it is an attestation. You have to work with the AICPA to get licensed to or accredited in order to offer these reports. And then yes, the badge. I don't remember exactly what the badge looks like, but you know that that's that's what you're delivering. David Leary: [00:18:52] And so, so if we want to start, I'm assuming the way this would work, you and I want to start up a new company tomorrow and we're going to do soc2 stuff, right? We probably have to go apply somewhere from the AICPA website, somewhere to be allowed to issue these things and give out the badge. Blake Oliver: [00:19:10] We'd have to be a CPA firm, right? But here's the here's how you do it. Let's say you're a software startup and you don't want to do that. You can do all the work. David Leary: [00:19:18] The tech. Okay. Blake Oliver: [00:19:19] You do all the work, and then you just pass over this report to somebody who does have that accreditation with the AICPA to sign off. David Leary: [00:19:31] If the firm trusts you and maybe you wowed them with your tech or who knows what. They might just sign off. Or if they do it. Once you investigate it once in the first one, then you assume all the future ones are in compliance. Blake Oliver: [00:19:44] And I mean, I think that might be even generous, David, because like, what if you just, you know, get a, you spin up a CPA firm, you get the accreditation, and then you just issue these soc2 badges, you sign off on these reports, even though you haven't actually done the work. David Leary: [00:20:00] So just like for us, like we want to do a CPE course with Nasba, in order for us to use that Nasba logo, we have to prove that we're meeting the standards and requirements and we and we can be audited at any time randomly, right? Yes. So does this. Is the Icpa checking on all these badges that are on everybody's websites? Blake Oliver: [00:20:19] I don't know, I would love to know. David Leary: [00:20:23] David go out to their website. I noticed they don't have any accounting jobs on the website. It doesn't mean they don't already have a CPA hired on staff. I don't know, but they're not hiring for any. And I thought it was interesting. I was drilled down on the founders LinkedIn see what they're posting about. They actually just did a webinar three days ago called Compliance is Proof that You Take Security Seriously. That was the webinar from three days ago. Blake Oliver: [00:20:49] David, let's go ahead and thank our next sponsor of this episode. And that is Cloud Accountant staffing. Are you tired of the endless search for qualified accounting talent? You're not alone. Growing accounting firms are struggling to find available and affordable team members when they need them most. And cloud accountant staffing has the solution with their revolutionary candidate portal. Unlike traditional staffing agencies that waste your time with sales calls, paperwork, and deposits, the Cloud Accountant Staffing Candidate portal gives you instant access to highly vetted, qualified accounting professionals. While other firms make you wait weeks or months, Cloud accountant Staffing delivers candidates in just weeks, and their boutique support ensures you're getting quality talent that's both available and affordable. No waiting, no hassle, just top talent right now. The cloud account and staffing candidate portal puts you in control. Browse live candidates, make selections on your timeline, and build your offshore team without the traditional headaches. To find, review and book interviews with potential team members, all in less than ten minutes. Head over to The Accounting Podcast C a s, that's The Accounting Podcast dot promo forward slash c a s. Now. David, you had a story that intrigued me or a statement you made before we started recording. And it was maybe no one is using AI for accounting. David Leary: [00:22:11] Yeah. So this is, this seems a little crazy, but I'll put the link down in the chat. So all our live stream viewers and you can actually go to it as well. Give me a second here. Blake Oliver: [00:22:21] Oh, right. And I need to put the link to that Substack story in as well. I'm going to do that. David Leary: [00:22:27] So anthropic did a survey global survey of 80,000 people. Blake. And they took quotes from the survey. I'm assuming maybe the survey could have been online. You know, like you get that little extra box to add notes to a survey. 80, 80,000 or 81,000 people were surveyed. I searched for the word accounting. Only one quote had the word accounting in it. So I'll read this out loud. I work in accounting at a shipping company, but management asked me to build an automated profitability and an analyst reports program. With cloud help, I completed everything from the design to the code. I got promoted to manager and was selected as an outstanding employee with accommodation. This is a financial professional in South Korea. I searched for the word bookkeeping. Zero results. The word Excel one quote. Right. Uh, I searched for the word tax. Zero result. The only thing that came up tax wise was A. I'll save that for after this a little bit, but it wasn't taxed in the way we talk about tax. Right. So you figure if there's 80,000 people that were surveyed about AI and they're asking them about their how they're using it, their hopes, their concerns around AI, you would think words like accounting and tax and bookkeeping would should show up in some of these quotes. It's just nobody's using it for this stuff. Or if they are, they're super happy and they don't have any concerns. I don't where's the quote that like hallucinations mess up my financials? Like there's no quotes like that. I'm shocked actually, that there's no there's not more quotes regarding accounting bookkeeping, tax. Blake Oliver: [00:24:02] I'm glad you brought this up, David, because I have a story about how people are actually using AI for tax. This was in the Wall Street Journal and it's called Wall Street Journal. Readers share how they are using AI for tax prep. From capital gains to Roth conversions, people reveal how artificial intelligence is helping them and where it's falling short. So let's go to the examples. Brian Segnet he used ChatGPT to well, he tested it to see if it could take a volunteer tax preparer test. He's a retired Xerox marketing manager and he is a volunteer that helps. He helps people with their taxes. And he gave it, uh, an exam like an IRS exam that you have to take to be a volunteer tax preparer. It is for the AARP tax aide volunteer program. You have to pass an IRS test in two tries. And he says that when he uploaded the test into ChatGPT as an experiment, the chatbot failed for the second attempt. Segnet tried the test on his own, and he scored a 97, and as a result, he said, quote, I would not want AI to do even a simple return, unquote. Now I have my thoughts on this. I, I want to know exactly how he did this test, because we've been hearing for a while now that AI can pass these types of exams. I mean, even the CPA exam. But I think maybe what he did was try to one shot it. So if you just take the whole exam and you give it to ChatGPT, it's not going to do as well as if you give it one question at a time. And, um, also like you have to give it all the supporting materials. You have to give it reference. This is an open book exam. So if you just give it the exam and you don't give it all the supporting material to hypnotize. David Leary: [00:25:58] It, you have to say you're going to be a tax professional or a tax preparer. You have to hypnotize it first, then give it data, then ask it each question. Blake Oliver: [00:26:07] So you know, I'm not I'm not totally sold on that one, but that wasn't really an example. Let's go on to more of the, uh, more of the actual examples of how people are using it. So the next reader is Keith Squires. He's in Attica, Ohio, and he used Microsoft Copilot to model Roth conversion timing. He created a spreadsheet to estimate the tax cost of moving money from a traditional IRA to a Roth IRA, and he was able to model the tax consequences of these conversions and then used market dips to convert when the asset values were temporarily lower. So the future rebound could happen inside the Roth. So it's more than just planning, right? It's helping with scenario modeling and spreadsheet creation. It's a good example of AI assisting with tax planning and not just like filling out a form, which is like a really interesting use case because, you know, that's a, that's the thing that, uh, is really expensive. You want to hire a CPA to do that. David Leary: [00:27:15] It feels like the two examples you just gave me, though, are like experts in their field attempting to use AI to do like this Roth conversion thing is like way high level. Like when I see when you came with this headline, it says Wall Street Journal readers, I'm assuming it's the common man. How is the common man using AI for tax prep? Well, oh, I, I designed this. Blake Oliver: [00:27:35] Whole. David Leary: [00:27:35] System for a Roth IRA or Roth conversion. Nobody's just doing that or rando. Blake Oliver: [00:27:40] Well, to be fair, right? Wall Street Journal readers might not be okay. David Leary: [00:27:45] Excuse a little high. Blake Oliver: [00:27:46] But neither neither are the listeners of our show. Right? Um, I mean, hey, if you're just, you know, an average person doing your tax return, I mean, one of the best ways you could actually use AI would be like, use the Chrome browser or use Microsoft Edge. And then as you're going through your tax prep software, you can ask all your questions to Gemini in the sidebar or to copilot in the sidebar, and it can see what's on your screen and help you through it in a way that like the tax prep software struggles because they're using, you know, older, uh, tech, right? Just like help dialogs and, and, and help articles, whereas you can get like specific answers to questions. I mean, that would be, that would be my use case. Actually, that's what actually I've done. And it works. It works really well. Here's another. David Leary: [00:28:33] Example. Blake Oliver: [00:28:34] Oh, go for it, David. David Leary: [00:28:36] This just to reinforce here, um, this is in Twitter. Um, somebody texted me this. That's I had to get off my phone. There's, uh, what is it? This is from March 19th. I'll read the tweet. Grock just saved my sister $1,441 on her taxes. I had it, checked the TurboTax output and found it. Found a mistake. Seriously, uh, 4.20 is very good with taxes. And then it got retweeted by Elon Musk that says says try using grock for your taxes. So I think a lot of the common use cases are going to be way lower than this with this article is going to have like just here's the output, review it. Blake Oliver: [00:29:15] Here's a more day to day example. Maybe you like this one. David. This is Cindy Piercey from Mount Pleasant, South Carolina. Cindy used Gemini to. Well, what she did is she gave it a list of donated household items and it generated a spreadsheet with estimates of value. She reviewed the output and then used that to value her charitable contribution for donations at thrift shops. But there was an exception for a 1904 complete Edgar Allen Post-It she had to use eBay sales instead. It didn't go and get the the value of that. But again, that's the sort of thing where you might have to give it more context. Maybe you need to go one item at a time to do this, because Gemini could certainly go out and do the research and maybe even find comparable sales on eBay, but it's not. David Leary: [00:30:08] Going to maybe go use eBay. Go. Yeah, you have to give more granular instructions. Blake Oliver: [00:30:12] It's not going to go one shot that if you give it a spreadsheet all at once. Right? It's going to it might you might with a chat bot like that, you might have to prompt it more specifically multiple times. Go step by step by step. Although one of these new tools like clockwork, which can then take a task and break it down into multiple steps, maybe it could figure out how to do this without you having to go and tell it to do that. So you know this, this could be a great way to accelerate your documentation. Figure out valuation for charitable contributions, stuff that's rare, maybe market specific, it won't be as good at. But I mean, that's a big time saver. Um, good example. David Leary: [00:30:54] Especially if the client just comes the list of stuff and they just give it to you. Where do you start? This would be easy for you to get it close and then you go out to the client. These values seem right. Yeah. Blake Oliver: [00:31:05] Jorge Mariano's of Murray, Kentucky used Microsoft copilot to understand an IRS letter. He got a confusing IRS notice and then he used AI as an interpreter. So he uploaded the notice he got from the IRS. It was about an unexplained 1099 int for about $155, and the AI was able to explain that the interest was likely paid because his refund was sent more than 45 days after filing. So that might be one of the most practical use cases, and I've used it for this before, is helping regular taxpayers. And I mean, this could be helpful for staff accountants, newer tax people who are not familiar with these notices to understand them and figure out what's going on and how to respond. David Leary: [00:32:00] Or IRS should just use AI and say, create a cover letter that explains this document I'm emailing. I'm sending somebody the tax payer like a five year old. Explain what this this notification is. I'm sending. Yeah. Blake Oliver: [00:32:16] Steve Kliner of Lakemont, Pennsylvania used AI to do data wrangling. His source material. It was semi-structured text and documents, and he used AI to convert it into cleaner tables that could be used for tax work. So he and his wife, they sell home decor on eBay, and his wife writes item descriptions in Google Docs and includes cost details in their. He ran that through AI to turn it into tabular data and said it was a huge time saver. So another great example of something AI is really good at. It's taking that unstructured data in a document and turning it into structured data that you could then use to get the number that you need to put into your return. David Leary: [00:33:06] And you know, somebody just sending those Google Docs as is right, to their accountant, right? For sure. Blake Oliver: [00:33:12] James Dillard of Franklin, Tennessee, used grok to estimate capital gains tax and net investment income tax on stock sales. He had a large Apple stock sale in 2025, and he tried using AI to do a technical estimate involving capital gains and n I I t. And he said the first answer was wrong because AI missed a key bracket issue. Only after he changed how he presented the problem did the tool get it right. So David, that goes to your point that with some more complex stuff, it might not be there yet. David Leary: [00:33:53] And you know what that's actually called. And we didn't cover this from my anthropic quotes. That's called the fact check tax. The what the fact check tax. Okay. So as you know, and I'll read the full quote where this was used in an assistant that sounds sure, but it's often wrong. Forces you to treat everything as suspect. Instead of freeing attention, AI creates a permanent fact check tax. Right? So. And that takes brain power. Yeah. Fact check tax. That's my new favorite AI related term. Blake Oliver: [00:34:26] Paul of Boston. Use ChatGPT to estimate tax payments related to a Roth conversion and investment income, mainly as a way to avoid doing the calculations manually or contacting his accountant. David Leary: [00:34:39] That's that's the real motivation. Blake Oliver: [00:34:42] And finally, Glenn Daly of New York used AI to compare tax burdens in foreign countries for retirement planning. He was seeking to understand international tax consequences for possible expatriate or extended stay planning. That could be the most dangerous use case because we know cross-border tax. It's it's very fact specific and residency rules. They get complicated. Very complicated. Really fast. He was looking at Canada, France, Mexico and Spain, and he found that AI suggested income taxes for seniors were much higher in those countries. I'm not sure. I guess we don't know if it was right or not, but that was an example. So what's the takeaway here? Across all these examples, AI was strongest at support work, not the final judgment. It saves time. David Leary: [00:35:44] Takeaway is there avoiding getting an accountant or a tax professional person? That's a takeaway. Blake Oliver: [00:35:51] Well, yeah, that's the this is the sort of stuff you might go to a tax pro first for. Right? And now people are going to AI first. And it's definitely not firing on all four cylinders. So will this end up just creating more work for tax pros? I mean, that's been the case in the past. All new technologies automate something but create more messes. We saw that with bookkeeping software, right? Quickbooks comes out. Suddenly people are thinking, I can do my own bookkeeping. I don't need. David Leary: [00:36:22] To write a check. Blake Oliver: [00:36:23] If you could, there's more. David Leary: [00:36:24] Work you could use QuickBooks. Blake Oliver: [00:36:25] Yes, there's more work for bookkeepers and accountants than ever before. So maybe that's the lesson of this article. David Leary: [00:36:33] Did you want to read our next ad? And I'll get ready for the next story. So I have a story about, uh, work games, and then I have a game for you and I to play. Blake Oliver: [00:36:41] Are you tired of payroll headaches getting in the way of client experience? The client experience you want to deliver manual workflows, creating bottlenecks, compliance nightmares, and endless support calls that go nowhere. There's a better way for your team and your clients on pay. Is the payroll partner that accountants and bookkeepers love. Why? Because it's easy to use, packed with value and backed by support that actually supports you. Their team gets rave reviews for being fast, expert and reachable when you need them. Onp handles the heavy lifting. You get a dedicated onboarding coordinator who sets up worker profiles and transfers year to date data from previous providers, all at no extra cost. Their seamless QuickBooks and Xero integrations eliminate manual journal entries, and they support any type of business you serve. Farms, restaurants, nonprofits, you name it on pay can handle unique requirements without adding complexity on pay. Keeps pricing simple, too. Everything your clients expect from multi-state filing to off cycle pay runs is included. No hidden fees, no surprises. To book a demo, head over to The Accounting Podcast. That's The Accounting Podcast dot promo forward slash o p a y. David Leary: [00:37:53] So, Blake, back when you were at the firm, you worked for a couple big firms and you owned your own firm. Did you guys ever play games at work? Blake Oliver: [00:38:01] I, I know we did some stuff like that, but mainly March Madness brackets. But yeah, that was that was it. I think I did it one year, but I'm, I'm no good at that. But apparently it doesn't matter. So. David Leary: [00:38:14] Well, something you might want to do at your firm to help with busy season stress. So there's some some benefits to games. And I'll give you ten ideas for games you can do. So you get productivity games, games that are trivia or competitive. Um, you get performance productivity boost between 14 to 30%. If you're doing those at your firm, 14. Blake Oliver: [00:38:32] To 30% productivity boost from doing. David Leary: [00:38:35] Because if you do trivia games at your work, if you do physical activities like have a sports day, productivity can boost up to 21% in your firm. Blake Oliver: [00:38:43] Um, can I stop you for a second and just ask you like, where is this coming from? David Leary: [00:38:47] Oh, this is small business trends.com. So that's an American express, uh, blog, right? If I remember correctly, I think that's American Express's website, small business trends.com. I think, okay, don't, Don't fact check me on these. Okay. Um, but when you do, when you play board games at work, you have board games available. Uh, job satisfaction increases 20%. Absenteeism decreases 25%. So sometimes that board game is a reason somebody coming to work. So it's important to have these, um, brain teasers, puzzles like that. You have 21% in productivity, which you just talked about, but also improve team communication. And this is the one I think is really funny for accounting firms to do escape rooms. So escape rooms can increase your team cohesiveness by 40%. Blake Oliver: [00:39:38] 40% doing an. David Leary: [00:39:39] Escape room isn't like working at an accounting firm during the busy season, just the escape room already. Like you have to solve all these puzzles or you don't get to go home, like, right? Blake Oliver: [00:39:50] Like, yeah, right. David Leary: [00:39:51] What's the difference between an escape room and busy season at a firm? Blake Oliver: [00:39:55] A busy season at a traditional firm. That's funny. David Leary: [00:39:58] Yeah. So here's the ten, ten games you could do. Show and tell sessions. You could do board games. You could do workplace contests, puzzles or brain teasers. Office scavenger hunts. That one seems fun. Trivia games and employee sports games. Escape rooms. Themed office days, including dress up decorations and snack themes, and team volunteering activities. All of these really take the edge off of work and help build your make your team better, ultimately. And with that. Blake Oliver: [00:40:26] Data, I mean like this. This is, this is like somebody studied this and figured out it actually works. It's not just. David Leary: [00:40:33] On the article and see it. Do they actually quote, who's the expert on this? Blake Oliver: [00:40:36] Where did the where did the data come from? Yeah, I'm always going to ask that question, David. You got to know. Well, while you look that up, I want to talk about a threat to the CPA profession that was narrowly averted this year. The Florida Institute of CPAs says it successfully helped stop legislation in the 2026 Florida regular session in the legislature there that would have eliminated the Florida Board of Accountancy. They published this on their site, and as part of an advocacy update from the CPA, there was a bill in Florida called HB 607 Industries and Professional Activities, and it moved quickly through two two committees. And the bill did not clear the final committee and did not reach the House floor, which prevented it from becoming law. And it it it would have eliminated the Board of Accountancy, along with a bunch of other professional licensing boards, as part of a deregulation effort in the state of Florida. David Leary: [00:41:45] So no more cosmetology licenses, no more real estate agent licenses. They're just like a broad hammer type situation. Blake Oliver: [00:41:52] Yes. Um, it was the most serious according to the Florida Institute of CPAs. It was the most serious threat to the profession in decades. And I have to agree. You eliminate the Board of Accountancy. I mean, how do you regulate the CPA in Florida? Unfortunately, in order to stop this bill, the Florida Institute of CPAs had to table its own effort to create a streamlined licensure process, which is, in other words, the alternative pathway that would allow for 120 hours or not 150 hours of education, which is too bad, because Florida was the first state to put the 150 hour rule in effect. And so I'm really looking forward to when they manage to undo that in the future. David Leary: [00:42:45] 30 states now, I think. I want to say Idaho just approved it. So we're at 30 states now that have alternative pathways. Blake Oliver: [00:42:52] So they're going to retry this. They're going to refile their licensure bill in 2027. And they're going to continue to advocate for new licensure pathways and efficiency reforms for future sessions. So it's a good example of the of, you know, your state CPA society working behind the scenes to protect the CPA. So congrats to Fi, CPA and the leadership there for protecting the CPA license. It's tricky because we want to streamline it, but we also don't want it to go away at the same time. So you've got to it's sort of like fighting battles on both sides. We've got the folks that want too much regulation, and then we've got the folks that want no regulation, and there's got to be a middle ground here. David Leary: [00:43:41] I imagine with the AI companies and the money, we're probably going to see more pushes for this because people are going to want the big, huge AI companies are probably going to want to have their AI, do CPA work and not have a license be in the way. I just I just imagine the funding, their lobbying efforts for these legal initiatives. This won't be the first or the last. We'll see. Blake Oliver: [00:44:02] And now let's think our last sponsor of this episode, and that is u n c. Let's face it, the job market is especially tough right now. But every industry needs accountants and accountants are always in demand. In fact, employment for accountants is projected to grow 10% through 2026 faster than most other professions. That's where uncW Kenan-flagler Master of Accounting program comes in. It's one of the top ranked Macc programs in the country, with 98% of students accepting a job offer within three months of graduation and earning more than those with just a bachelor's degree. If you're currently working full time, raising children, serving in the armed forces, or living halfway around the world, they're highly flexible. Macc program can also fit your lifestyle. You can choose their 12 month on campus program or their online only option, where you have up to 36 months to complete your degree. Plus, you'll join the powerful of 46,000 strong. Unc Kenan-flagler alumni network connections that will serve you throughout your career. If you want to set yourself up for lifelong success, pick the Macc program with proven ROI. To see why you should get your Master's in Accounting at the UNC Kenan-flagler Business School, head over to The Accounting Podcast. That's The Accounting Podcast 1099 u n. David Leary: [00:45:20] All right, Blake, are you ready to play a game? Blake Oliver: [00:45:23] Hit me. David Leary: [00:45:24] Hit you. Okay. So I got an ad, very targeted ad on LinkedIn. And I want you to guess the app, what the app this is. So? So I'm going to just refer to it as X and I'm going to read and it's one of those, you know, like on LinkedIn where you get those ads where it's like a slider of almost like slides and they come across there's image after image after image after image, and then at the end, it's like, sign up today or something like that. So I'm going to read you the value props of this ad and the the images. And then you tell me what app this is. Blake Oliver: [00:45:56] Well, let's call it y instead of x because there's already an X now. David Leary: [00:45:59] Oh that's true. Okay, so y. Blake Oliver: [00:46:02] Or maybe let's do Z. David Leary: [00:46:04] Or z. All right. Z. Z handles your recurring financial workflows. Consolidation reporting analysis. So you can dedicate more time to insights that shape company direction. You handle the financial strategy. Z handles the rest, organizes your receipts into a clean spreadsheet. Makes more time for numbers that matter. Build the quarterly revenue models from your assumptions. Cross-references your docs against the checklist for a month end. Close. Try Z for finance. So what is Z? What is it? Blake Oliver: [00:46:38] Well, that sounds a lot like QuickBooks. David Leary: [00:46:41] Sounds a lot like QuickBooks. Okay. I don't know, is that your guess? Blake Oliver: [00:46:45] Final answer. Let's say final answer. Yeah. David Leary: [00:46:47] Believe it or not, it's an ad for Claude Kovac for finance. Blake Oliver: [00:46:50] Really? David Leary: [00:46:52] Not kidding. Okay, so so if you think about it, third party app developers and accountants and CPAs that are using cloud essentially train the model so they could just take everybody out of the middle. Quickbooks is now the fact that you thought this was QuickBooks Add is amazing to me. Blake Oliver: [00:47:09] I mean. David Leary: [00:47:09] It just. Blake Oliver: [00:47:10] Shows that what they've been talking about. Yeah. David Leary: [00:47:13] Yeah. They're going to replace the third. It's the history of platforms, right? Like if you go back to your iPhone when you first got your iPhone, the camera just took a photo. And now there's all these, you'd get a bunch of third party apps to do filters and editing the photos. It's just built into the iPhone now, right? That's what happens with these platforms. They might say they're not using your data, but they're watching what you're doing with the app. Do not use your data to train the model, but that doesn't mean they can't watch what you do and then create that as a separate product offering. So they have a product now separately spun up that's actually called cloud financial services. Now, what's interesting though. If you go to the website for that, what it says that app does on the website doesn't totally align to the ad. So maybe these ads are a test or a carrot to get accounting professionals to click through or finance professionals. But it was a very, very targeted ad that felt like a typical SaaS accounting product, actually. Blake Oliver: [00:48:09] Here's the website, Claude, for financial services. And I'm seeing today we're introducing a comprehensive solution for financial analysis that transforms how finance professionals analyze markets, conduct research, and make investment decisions with Claude. Is this what you're talking about? David Leary: [00:48:30] Yeah, this is the this is the actual the app. Yeah. Blake Oliver: [00:48:32] The financial analysis solution unifies your financial data from market feeds to internal data stored in platforms like Databricks and Snowflake into a single interface. Access your critical data sources with direct hyperlinks to source materials for instant verification, all in one platform with expanded capacity for demanding financial workloads. David Leary: [00:48:53] But that doesn't sound like the ad, does it? Blake Oliver: [00:48:55] No. Interesting. So. So it's like doing financial analysis. This to me is like more of a threat to like financial analysts than it is to accountants. But I see what you mean. Like they're, they're watching people use Claude for financial analysis, and then they're building the tool so that end users can do it on their own. David Leary: [00:49:25] I mean, what's next, Claude code tax or Claude. Blake Oliver: [00:49:28] Claude for tax preparation? David Leary: [00:49:30] Yeah. Claude. What's it called? Claude Cowork for tax. Yeah. Is that the next step? How many of these are going to be off the shelf just built in the model. And I think it's going to be a lot of these things. And that goes back to why did Intuit get in bed with anthropic and get in bed with OpenAI and build their own AI, because they have to bet that one of these are going to win and they have to, they have to hedge their bets against all these platforms. But I thought it was interesting. They're running ads. I mean, I get it and you know, I get ads, people, uh, I look like an accountant, right? And so I get a lot of ads targeting accounting firms and CPAs and tax professionals, you know, on all the like I also click the ads, which actually gets them. Then I get even more ads like this. But the fact that I got ads from one of the AI, what did you say? Like route technology companies, is that what we would call platform technology? Blake Oliver: [00:50:21] Platform AI companies? Yeah. David Leary: [00:50:23] Very specific ads that you think might have been a QuickBooks ad is really well. Blake Oliver: [00:50:29] And the only reason that I think that they won't be able to disrupt like a QuickBooks is because QuickBooks is more than just all the stuff that you do on top of the data. The real value of QuickBooks is it is the source of truth for the data. And as much as you hate or love it, uh, that's what it is. It's like the standard. And people trust it and they trust the data stored in it and the reporting from it. So it's not like these AI companies are going to go build a general ledger solution that people adopt. They're still going to have to rely on that data underneath. And so Intuit's going to keep those customers, and Xero is going to keep those customers. Unless somebody can build something that people will trust more. And that is where like Digits and puzzle are coming in to disrupt. But that's going to take time, like building trust like that. Or maybe I'm wrong. Maybe it'll just happen overnight and people will just move. But there's a big difference. David Leary: [00:51:27] Between because if, if you're, if your digits or puzzle and you're building this new AI GL thing, you should be scared to death that one day. Claude's just going to offer that as a service. Blake Oliver: [00:51:38] Well, I'll tell you what you could do right now with like Claude Cowork cloud Co-work is this desktop app currently that sits on your Mac, and it allows you to reference local files in a folder as part of a project. So you can create or you can have cloud create a spreadsheet and then work in that spreadsheet. And so it's not like you're storing all this tabular data, this database in the memory of the LLM, you've actually got a separate spreadsheet that is interacting with and that creates a lot more reliability. So because that spreadsheet is not probabilistic, right? That is deterministic. That is the data is here, it is this value 100% when these formulas sum up. And that is always going to be accurate because that's a mathematical operation, not a statistical one. So I think I'd actually like to try some of this as a test is, you know, there's all those business owners who keep their books just in Excel. I mean, you can do that, right? If you're cash basis, you can just keep a listing of income and expenses in a spreadsheet, and then you can run pivot tables essentially to get the information that you need. Do your taxes. And theoretically, you could even create a balance sheet. You could create like an entire set of like, you could create a GL in a spreadsheet system. It's not that complicated. It just has to link together properly. And so I wonder actually, if you could at this point use Claude Cowork to create that spreadsheet, general ledger system and have it just interact with the spreadsheet for you as the interface. And then you could do simple accounting for free, essentially the cost of cloud cowork. David Leary: [00:53:25] Which was my. Blake Oliver: [00:53:26] As your UI. David Leary: [00:53:27] Last week when I coded QuickBooks on accident. Right? Mhm. Yeah. Blake Oliver: [00:53:31] So, you know, I'm, I'm fascinated to see what's going to be possible given how fast tools like Cowork are improving. And how reliable the outputs are getting because of the way that it approaches. Doing a task, creating a checklist, going through the checklist. Madman Dan on YouTube says, I see the GL partners eventually embedding with partners. We have seen that the GL leaders like Intuit and Xero move very slow in adoption. I think maybe, maybe Madman Dan meant the AI partners, um, Intuit and Zero. I've moved very slow in adoption before you can say it. Um, now the speed of development is just too fast from the AI leaders. They simply can't keep up. And why bother? Who's going to touch Jack's or whatever Intuit has? That's true. I think like it's going to be really hard for like, Intuit and Xero to keep up unless they're just plugging into like ChatGPT or into cloud to do this stuff. How can their own AI chatbots keep up with what these companies are doing, how fast they're developing? So I think ultimately it's not going to be the AI chatbots in those products that people use. It's going to be the off the shelf ones. I have Claude and it plugs into my QuickBooks or my Xero data, or I have ChatGPT. David Leary: [00:54:51] Long term copilot. Like, why do I want a separate chat box in this app and a separate chat bot in this app, in a separate chat bot in this app. And I'm copying things from one chatbot to the other. Nobody wants that. You're going to have one Universal Assistant chatbot running on your computer. Blake Oliver: [00:55:07] I got one more story to take us out and this is a LinkedIn post of mine. I was reading the CPA journal and I saw some interesting stats about female partners in accounting. We've known forever that this has been a problem. Like gender diversity in accounting firms, especially at the partner level, has not historically been good because historically speaking. David Leary: [00:55:36] More women go in in the beginning, but getting to partner, that funnel just really collapses, right? Blake Oliver: [00:55:42] So here's where we stand, right? In 1989, the number of female partners and accounting firms was just 4%, 4% in 1989. It's 39% today. So that is a huge improvement. I mean, it took a long time to get there. I mean, what is that over, uh, 30 years? But hey, that's a big, big change. Still a long way to go though. Women make up 44% of the CPAs in this country, but only 26% of lead engagement partners on the S&P 500. So like a quarter, only a quarter of the lead engagement partners for the 500 biggest companies in the United States are women. So there's a there's a difference there, right? We've got 49% female partners, but only 26% get the biggest clients. And I posted that on LinkedIn. And in some sectors it's even worse. It's like energy, real estate, utilities, less than 20% female lead partners. And I just I posted that without any further commentary. David Leary: [00:56:58] And in theory, the ratio should be the same. All things equal, right? Blake Oliver: [00:57:03] Like, why wouldn't it be unless there was some sort of like glass ceiling or there was some sort of, you know, discrimination happening. And, you know, the comments, I just have to point out this one comment. One of one somebody commented. Tim commented and said, why is more women partners a win for the profession? And I, I just had to like take a moment and I responded and we actually went back and forth on this. I had to explain. I had to mansplain to him why having more women partners is a good thing for the profession. And I guess it, you know, it seems obvious, but maybe we have to spell it out for people. David Leary: [00:57:51] The fact that he asked that is almost proof y in a weird kind of way, right? It's like when people don't until they understand their incompetencies, they have the ability to understand that. They'll never understand that they're incompetent, right? It's kind of. Blake Oliver: [00:58:03] Well, here's how I put it, right? You know, I was trying to think like, how do I explain this? Like how, how do I, how do I, how do you it just seems like it should be obvious, but how do you explain it? So here's what I said. I said men and women, they think differently. Yes or no. David. Do men and women think different? David Leary: [00:58:26] Yes. I just had an incident this weekend at my house. Blake Oliver: [00:58:28] Okay. Does increasing the diversity of thought in leadership advance the profession? David Leary: [00:58:36] Yes. Blake Oliver: [00:58:37] So we have different viewpoints, right? That's better for the profession. Can we take that as a given? Yes. Well, then obviously increasing the diversity, the gender diversity in leadership in the profession is a good thing. It's good for the, the profession. And like, that's how I think about all of this stuff, right? Like diversity for diversity's sake, I agree. What's you know, why? I mean, that's a great thing to have, but like, hey, if you need me to spell it out for you, like this is why it's like people have different life experiences depending on whether they're a man or a woman or if they're white or they're black or whatever it is. Right? Like it's a different perspective. And so if you only have people at the top that have one perspective, you're going to get less diversity of thought, less diversity of experience, different thoughts. Like that's why I think the accounting profession has been so stodgy for so long, because it's all these people with the same experience sitting there at the top who don't. David Leary: [00:59:47] Like group groupthink. Blake Oliver: [00:59:48] New ideas. Right? And also, there's another great argument for it, which is you've got all these people working for your firm that are women. You've got all these clients that are women. Don't you think it would be good for the firm if, like the leadership reflected the people that work for you and the clients, you probably make more money. David Leary: [01:00:14] Yeah. Blake Oliver: [01:00:14] So like, if that's all you care about, you still have a reason to do it. You're going to have more engaged employees and more engaged clients because the people running the place look like them. David Leary: [01:00:29] And historically speaking, you're probably paying all these females less. So now your firm's even more profitable if you want to go to that whole other problem. Blake Oliver: [01:00:37] That's terrible. David Leary: [01:00:38] That's what I'm trying to say. Like if all you care about is the money, right? Yeah, that would be that would be logical for you to think that way. But people aren't going to think that way. I'm not saying it's okay to pay less, but. Blake Oliver: [01:00:49] Thanks everyone who joined us live. Great to have you with us this hour. Just like blew by David. David Leary: [01:00:54] It flew. Blake Oliver: [01:00:55] It flew. So much fun. If you want to earn free continuing professional education for listening to this episode, watching this episode and all of our past episodes, go to earmark app or download the free app from the App Store. Create your free account and earn one CPE for free every week. Subscribe for the low price of $170 a year at the moment, and get unlimited CPE every week and access to premium content. David Leary: [01:01:22] And you'll get Nasirb approved CPE. But you actually have to do the quiz and earn the CPE. This is not just a certificate factory here. Blake Oliver: [01:01:29] No, you've got to pass that quiz. Yeah, we we did the hard work. We actually are doing the real compliance. So, uh, and yeah, let's let's I wonder, you know, I wonder how often nasba like investigates, like CPE providers like, you know, are there, are there providers out there just rubber stamping certificates and issuing them? Like I wouldn't audit it a couple times already. David Leary: [01:01:53] Correct. Blake Oliver: [01:01:54] Oh, yeah. We've been audited multiple times and we've passed. So, uh, it's hopefully they're doing that, uh, with, with all the providers and not just us. All right. Thanks everyone for joining us. We will see you here again next week. Uh, and, uh, stay, uh, stay warm. If you're in the Midwest, stay cool. If you're here in Arizona, stay safe if you're flying and see you next week. Bye, everyone.