
Loading summary
A
Welcome to the AICPA Town Hall Series, your resource for the latest news and updates on pressing issues facing the accounting profession.
B
Good afternoon and welcome to the AICPA Town Hall. I'm Eric Ouskerson, one of your hosts for today. Today is January 22, 2026. Mark and Lisa, our second town hall of the year. It's great to be with you both. I'm coming to you live from New York. We got Mark PETERSON in our D.C. studio and Lisa Simpson in Durham. Mark, we got and Lisa, we have plenty to talk about, but we're also talking about the weather. You know, right now people are talking about a historic snowfall potentially, and I don't know historic, but the biggest snowfall in the last five years coming to New York. And I know Washington, D.C. is thinking about it. And Lisa, it sounds like you got an ice storm forecasted in Durham.
C
I'm just hoping I don't see Jim Cantore anywhere in the Raleigh, Durham area. And I know we'll be okay.
B
Well, we've got a great show. We're going to kick things off. Mark Peterson is going to give us an update on what's going on in Davos as Well as Washington, D.C. so it's not just a D.C. update, it's D.C. and Davos. I've got a great session. We're going to be talking about some AI resources that we're bringing to the membership via CPA.com, and we're going to be having a discussion with Jeff Siebert, who's the CEO of Digits, on those resources and really the state of AI. Lisa, I know you've got an interesting segment on digital asset income tax reporting.
C
Yeah. I think it'll be a new topic for a lot of us and I'm lucky to have a couple of great folks that will walk us through what we can expect.
B
Okay. And we've got Melanie Lauritson, who's going to be giving us an update on many things going on at the irs. And we also have an IRS poll that we'll be giving you. But let's kick things off right now and looking at Snowy Davos.
D
Mark, Snowy Davos. Maybe more snow here than in Davos, Eric. We'll find out. So let me just get some overview of thoughts. I will say this, Erik, from my perspective at every level of the profession, I have never seen geopolitics have a bigger impact or potential impact across the profession on different issues. And if you look at the platform, the economic platform that the president has been discussing and his cabinet members that are in Davos discussing its tax, its tariffs both for revenue driving manufacturing back to the US and then geopolitical leverage. And we're going to talk about tariffs specifically in a second. But then deregulation is a major theme as well as immigration. And you think of the impact on students coming to the US to get into accounting and then also high skilled visas. So there's a lot going on in that agenda that actually impacts multiple levels across the profession. But you can't talk about that platform without also thinking about the reality of the politics, which is we're headed into the midterm elections and affordability and I'll talk about that in a second as well, is a significant issue going into those midterm elections. And so when you think about what we're going to be watching is the redistricting that's going on in several states, Texas, California, Virginia has just announced several other states as well could change the dynamics. We've got. The retirements are starting to tick up that are being announced. But then there's also resignations. There are members of Congress that are leaving before the end of their term. And with such absolutely narrow, hair thin majorities, you know, someone leaving early and meaning a seat is vacant for a month or two until the governor decides how they're going to replace that member can have a huge impact. And we actually even saw that earlier today as they were working on some funding stuff. But if you think about that affordability narrative that the White House is pushing because they're trying to figure out how to equip Republicans in this case in order to show well in the midterms because it is a significant issue that's coming through in the polling. They're focusing on housing. We've heard about capping credit card rates, energy. And also there's a dynamic around tariffs where in one way tariffs are being utilized for revenue. They're being utilized for a lot of geopolitical leverage, but they also have a impact on the cost of goods. And so we saw adjustments already earlier this year to some of the tariffs related to some commodities, things that are on your breakfast table. But how does that all work? Are things that are we're going to be rolling out seeing and watch rollout now. The first primary is in March, Eric, so it's it's game time. This is all happening right now.
B
Well, Mark, if Mark Kozil says this all the time, if there's ever time not to just be chasing every headline, it's right now. It's been a very Busy week. A lot of, lot of swings.
D
Oh yeah.
B
Advisors just kind of, you know, staying steady and kind of, you know, maybe letting things just kind of sort themselves out without, you know, you know, jumping at, you know, every, every news announcement.
D
Well, let's talk about the tariffs here. And so, you know, obviously going into the, to the, into Davos, there was a lot of discussion that had popped over over Greenland and and U.S. trump administration aspirations there. Tariffs came up as a way to push back against opposition to that that was announced, that has since been retracted. But there was an immediate reaction of the EU suspending their affirmation of the trade deal, the EU US Trade deal that will probably get rolling again since those retaliatory tariffs have been pulled back on. And obviously there was a breath of relief when the idea of a aggressive posture towards Greenland has been moderated. But when we think about kind of the action reaction, I mentioned that affordability platform going into the midterms. But then again, we've got all of these tariff deals that are still in negotiations. Some of them are close to being finalized just like the eu. It just does extend this kind of period of uncertainty. One of the things that was discussed, and this is, you know, we're always looking through this as a lens of the profession was the EU has the ability to implement anti coercion tools and they could have done that and focused on services. Currently US Has a service surplus, trade deficit, but service surplus. And that directly hits the profession. Now hopefully the temperature has calmed over the last day and we will not be moving forward on that. But hearing services being mentioned in this debate is definitely something that the profession has to keep an eye on. We've got other things that are going to hit us this summer. They're going to be renegotiating the U.S. canada, Canada and Mexico deal. That is another one where there's a lot of tension around it and so very important again to the United States. We'll see how that plays out. All of this will be again put through the filter of those midterms.
B
Moving on to appropriations. And what's going to happen with the whole, you know, the budget and government fund. Are we going to have another shutdown?
D
No, we're not. That's my prediction. So we'll see if I'm right. The government funding runs out on Friday next week. They are, they actually voted, it was 214 to 213 in order to allow for a rule for debate of four of the appropriations bills next week. We think they're going to continue to roll out those bills to get them with full years of funding, including one, and Melanie will talk about this, that that will fund the irs. There are some, you know, there's some controversy related to the Department of Homeland Security because of controversies around ice. There could be a holdup of there. But we think that they'll probably get the bills they can get done for full funding for the year they will get done. And anything that's left over, any funding that isn't agreed to, they will put in a continuing resolution and then kick that that can down the road to probably after the midterm elections in order to finalize the funding. So we don't anticipate a shutdown, although President Trump was quoted earlier today saying it could happen over Department of Homeland Security funding. But I don't see an indication that the lawmakers are interested in not getting it funding. And they're also trying to get out of town here before the snow hits.
B
Maybe that's one of the positives in the storm.
D
It is a positive. And then the crypto discussion has been ongoing. And we've been talking a lot about the Genius act, which really set up the infrastructure for stablecoin. There is a second package called the Clarity act. And in it they really do the market structure around that, between the CFTC related to commodities and the SEC related to securities and kind of that framework of regulation around that. There's some cleanup language in that legislation that we're very interested in that would give give us more direction around some of the controls related to financial statement audits in this space that was scheduled for for this week. Ultimately, it got pulled because of some controversy around debates around yields and ethics about engagement of members of Congress or the administration officials and crypto. And so how this will play out is kind of yet to be determined. We had one of the major actors in Coinbase who flipped to opposition because they were concerned about some of the elements of the bill. It is definitely an important part of that debate, but there still are some challenges. So we're going to be watching again to see when they come back next week if they can make some progress, although I think that this has really harmed the chances of getting a quick resolution on that. Eric?
B
Yeah, Well, a lot happening the Genius act and now new bills addressing similar topics.
D
And then another issue that we've been talking about because it's so important as it relates to debt and deficit and fiscal responsibility, which was just introduced reintroduced. We've had this introduced. Working with our supporters on Capitol Hill in a bipartisan way that would require the House and Senate budget committees to come together together be required through a resolution to go through the financial statements of the federal government in order to really focus on where we are in our debt situation. And you know, you can we come to the brink close on defaulting on the debt and we are statutory required to vote those limits up. We've done that. But there, there. And this is a discussion that's going on at Davos. It's a discussion that's going on around the world, but specifically in the US of you know, what would happen related to inflation, you know, what would we, what would we do related to, you know, the cutting of spending. And so the idea is to have some thought leadership around this prior to when we get to an actual cliff or a crisis. But it is, it is a growing concern for policymakers.
B
Yeah. And it's something that the aicpa you've been active in, you know, for a decade plus, Mark, this topic, we've been.
D
There and again there's bipartisan interest but it's just tough to get stuff done. And then this is an issue that we've talked about a little bit, but I want to bring it back because we' to continue to be engaged in HR1 they was the language big tax bill. There was language related to some reforms to higher education funding. And as that was implemented by the Department of Education following Thanksgiving last year, they utilized basically two buckets to put caps in one bucket considered professional, which had a higher cap of $50,000 annually for student loans. The other one non professional 20,500 they put accounting engineers and architects in a non professional. Obviously that is a significant issue for us. We have already commented. Mark Cloziel has made statements. We're working with all of our partners with the state societies, others in the accounting community, the internal auditors, the other associations that focus on accounting. Basically if you combine membership about a million members in order to push back on how they want to do this. We are noted in statutes from the 33 and 34 securities Acts, Sarbanes Oxley requirements around professional standards. We're obviously a profession regardless of how the Department of Education has outlined this in order to get to the caps. But we think that this should be addressed. The comment period has not opened. When it does open, we're going to be soliciting memberships, the membership town hall community and state societies to comment. So more to come on that want to give everybody kind of the latest to play.
B
All right, Mark? Well, a lot of questions have come in we'll try to get to some of them in open forum and there's obviously a lot of support here for what you're doing related to pushing back on this. Accounting is a profession old degree issue. So. And as you've said, if we do want to do a call to action with the town hall, we definitely will activate them. But at this point, just stay tuned.
D
That's right. That's right. We're waiting for it to hit the register, Federal Register, that officially opens up the comment period and then we will be extremely active.
B
Thanks, Mark. Well, now we're going to move into an important topic. We talked a lot at the end of last year on how the technology trigger of AI, where things stand with AI are really driving some significant productivity gains and opportunities for the profession. But it is complex. And what we're going to do here is unpack the AI models and we're going to talk about some additional predictions. And I'm thrilled, I'm thrilled to have my friend and kind of technology partner here, Jeff Siebert, joining us. Jeff is the founder and CEO of Digits, which is an AI based gl. He formerly led consumer product at Twitter. The company he founded was acquired by Twitter undergrad from Stanford in computer science, very active with the angel community. And probably the final bullet here, which I really am glad to report on, is he's part of our AI working group. And@cpa.com, what we do is when you have these major technology triggers, we lean in and we do primary research. And what we're doing here with this AI working group is we've got technology leaders like Jeff, we've got firms and others that are kind of related to the profession. Pascal Fannette being one of them who are helping us really understand where things are going. Welcome, Jeff.
E
Eric, thanks so much for having me. It's great to be here.
B
Well, what we're going to first do is cover some items that we do outline in this toolkit. We've got this in the resource section. Jeff was one of the key authors related to this, breaking AI out into these five different components. Jeff, I'm going to let you do this. I do remember, you know, it's been since 2016 or 17 when we've been talking about machine learning and AI. So this has been going on for, for quite some time. So why don't you walk us through these five buckets?
E
Yeah, great. What we wanted to do here is share with the profession what the different types of AI models are and how they can be used what they're good at, what they're less good at, et cetera. Because as everyone knows, there's a ton of hype right now on AI. A lot of that is hard to understand. And every product in the world, every person in the world is like, oh, I'm using AI for X. Well, are you really? What is it doing? How does it work? And so to quickly run through these different types of models, we'll start with number one, generative AI. This is what's gotten all the buzz over the past few years since Jack came out in late 2022. And so when you hear the term LLM or Large Language Model or Gen AI, what you're talking about is a generative model and it's right there in the name. ChatGPT stands for a generative pre trained transformer. That's how they train the model. These models are trained on broadly all the written contents of the entire Internet. And so they have a vast amount of knowledge, but the way they work is they are basically generating tokens. So you give the model a prompt and let's say you go the quick brown fox, what is the model going to generate? It's going to generate the most likely completion of that prompt and so jumped over the lazy dog. These models are really, really powerful for content. You can write emails to clients, you can draft explanations, you could write an executive summary of your monthly financial report. But you need to remember that these models are fundamentally just generating text based on what they believe is the most likely completion based on how they were trained. So that's the first bucket of model. The next bucket is predictive models. These are a completely different technical architecture and, and you can think of these as categorization models. And they're really important for the profession, particularly in bookkeeping. And so if you look at bookkeeping, you are taking transactions, financial activity that's happened with clients, and you are booking it into the chart of accounts. You are basically classifying that transaction, right? You are categorizing into the chart of accounts in double entry, figuring out where it lands. The way predictive models work is they train on a set of outcomes and, and they give you the probabilities of each of those outcomes. And so you take the transaction. Was this travel and entertainment? Yes. That's high probability. Okay, it should go there. The important thing to note here is predictive models cannot hallucinate. So this is the technical term. Gen AI is prone to hallucination. It'll make something up. It can go a little bit crazy at times. Predictive models can't. And so in accounting that's very, very important. You don't want them generating, hallucinating a transaction, making up a new entry in your chart of accounts, et cetera, et cetera. Let's move on to document extraction. Everyone's familiar with OCR. There's been probably 20 years of OCR technology and tools people have used to extract invoices, receipts, checks, et cetera. That is all basically obsolete. So there's a whole class of document extraction models that use a technology called layout aware language models. And so not only are they understanding the text, they're understanding the physical layout on the page, on the document of how the information is presented. And that is how your brain's working. When you look at an invoice, you can see it's an invoice. When you look at a receipt, you know it's a receipt and where to look for the total, that's what these models are doing. And so they are now incredibly accurate at extracting data from documents. And so this is the future of all document extraction work.
B
And Jeff, before we go on to the last two kind of categories here, when you just, I mean this community, think about these, these first three, the first one, the generative AI ChatGPT, we're seeing a lot of opportunity in tax research being one category. So maybe comment on that. And then you look at predictive AI, that's something that I know you're leveraging a lot of digits. And then the document extraction AI, as you said, the next generation of this OCR technology, OCR technology really worked well. We're seeing it again in the tax area. There's a lot of startups in the tax area looking to kind of go the next level in, you know, automating the input of the, of the tech tax prep software.
E
That's exactly the right way to think about it, is when you're trying to achieve a task, think about how you or your team would go and complete that task. What type of thought process are you using? And so research is a fantastic one, right? You are trying to consume a whole ton of information about the tax code, about what's been said before, precedents, et cetera, and sort of condense that and summarize that into some guidance on what you want to do. The generative models are exceptional at that. When you're thinking about bookkeeping, that's more of a classification problem, like, where do I put this? That's a predictive model. And then for document extraction, of course, if you're trying to convert unstructured data into more structured data. That's where you would use those document extraction models.
B
Well, now the term of 2026 is agentic AI. So AI agents, yes.
E
This is where it gets really fun. So this is the cutting edge of the field and everyone's talking about agents. Most people don't really understand what they are or how they work. What an agent is is basically a large language model. So a gen AI that you give a harness to and you run in a loop. And so imagine you give this gen AI a prompt like, I want to achieve X. Oh, and by the way, here are all these tools you have access to. And those tools are traditional computer programs. So the AI could do some math, it could look something up, it could run an analysis. And so you prompt, the AI was saying, hey, what do you want to do first in order to complete this task? So the model comes up with an idea. Great. Goes and does that, runs the tools, gets a response, you then ask it, oh, did you finish the task? And it goes, probably not, I don't think so, I need more stuff. And it's like, great, what's the next step? And so this is where the loop comes in. And so you run this gen model in a loop until it believes it's completed the task and then it prints the answer. And what's really interesting is you can layer these models recursively and have them spawn sub agents that are doing subtasks and bringing their answers back together to create a very complex thought process, actually. And so agent orchestration is the hot topic in Silicon Valley right now. You can create some incredibly sophisticated workflows with this technology.
B
Well, let's move into the final category. Just one broad comment, and you and I have talked a lot about this. When I look at these technologies, the clients and the businesses really do want the firms to put these to use because there's complexity in configuring them, there's complexity in leveraging them. And Marcos said this many times, this isn't about automating the profession out of high value work. It's automating them out of the mundane, but allowing them really to kind of, you know, move up the value chain 100%.
E
That is entirely our mindset with digits is how can we use this technology to automate the tedium so you have the time to actually bring true value to the clients. And this is what your clients want as well. I mean, we speak with thousands of business owners and they don't want all the time going towards booking transactions and data entry. They want the actual advice and insight on how should they better guide and grow their business. And that's what technology like this really allows for.
B
So we got not actually AI. So it really is in our fifth category.
E
Yes. So this one is a bit of a lie. We wanted to call this out specifically because there's a lot of noise in the space that isn't true. And so, for example, let's say you're analyzing a P and L or a balance sheet, you're doing your flux analysis. There's a whole bunch of tools online that advertise, oh, you can do AIFP and A, and we'll analyze your P and L and tell you what to do. That is not AI. That is not a model. Computers have been really good at math for a very long time. And FP and A is largely statistics. Right. And getting the answers and then trying to figure out what to do with them. And that's where the advice comes in. And so be, be on the watch when you're talking with software vendors and they say, oh, our AI does xyz? Does it really? Or would it be way better if they just use some math to analyze the books for the month and see what changed? And so we wanted to call that out as a flag for everyone as you're speaking with folks.
B
Well, thanks. We're going to move into predictions now, but I just want for the town hall community, in your resources you have this document. It's an accounts primer on models and use cases. So it talks about a number of these use cases that Jeff just covered. And it also gives you six steps on how to evaluate AI solutions for your firm. Take into account, you know, these, these, these different models. So we're going to, we got two slides here in predictions. We're going to start with software engineering. You're a software engineer.
D
So.
B
Yes, by, by I was talking about college students. So what, what do they need to do? But here, by year end, you're predicting that 95% of software energy engineering will be automated. Like you'd expand on that and also to talk about what digits is doing. I mean, you're a technology company with a lot of software engineers.
E
Yeah. So we have two predictions to share. Both have really come about in the last 45 days, basically since early December. That is how fast this technology is moving. And the software engineering front, we wanted to start there because that is my life, my livelihood. I taught myself C when I was 12. I've spent my life programming computers and this is my passion and so to come to this realization is really just a shocking, I think, moment for Silicon Valley. And we have been acting very quickly at digits over the past three weeks to adapt to this. The reason I say this is there's a tool that Anthropic released this past fall called Claude Code. And their new model, Opus 4.5 is exceptionally good at writing code. And the big breakthrough they had is models have been okay at coding for a couple years now, but only in very simple code bases, like the most basic examples you can think of. And Opus 4. 5 excels even at large complex code bases. And that was the light bulb moment for me that realized the world is completely going to change in 2026 and we need to adapt to this. We did two things literally just in the past two weeks. We restructured our main code repository for the folks in the profession, think of that as your gl. Just like your ledger is the system of record for accounting, your source code repository is your system of record for your software engineering project. We restructured our primary repo at digits to be better suited for AI agents to help us work on. We also started a new code repository and a whole new project that is 95 plus percent agent written. And we have AI agents that have now output over a quarter million lines of code completely on their own, with some guidance and prompting into this repo. And the app works. It's very, very interesting. And so we are aggressively experimenting with this to prepare our team and sort of how we ship and deliver software to market for what now, I believe is inevitable over the course of 2026.
B
So what is the role of the future software engineer?
E
Yes, so it'll go much more into review and guidance. And so you still need software engineering. You need to understand how computers work, what they're good at, what they're bad at, all of the security implications and efficiency implications of delivering code to production. But you're going to be spending a lot less time on the tedium writing the inner loops and the switch statements and all of this stuff that just takes hours. All of these test cases you have to write and so on. And you're going to be focused at the higher level on what do I want my software to do, how should it be structured, what are the core components of it? That type of level of thinking, it's.
B
A lot more interesting. I remember writing Fortran when I was young. So your job's going to get more interesting. So now let's. I think that was a great analogy. I mean, it's something that's happening. Huge shift, huge, huge market. Here's another huge market that this town hall community is involved with. So month end close and on this town hall we got a lot of practitioners, Jeff. We also have about a third or over a third of our audience are in industry, they're in finance, in industry, you know, in control of chip jobs and things like that.
E
So yeah, this goes very much in line. And there's a famous quote from Alan Kay. The best way to predict the future is invent it. We share this because this is what is actively being worked on, not just by digits, but by a number of folks across the technology sphere. And it's the same exact mindset. The tedium of the month end close doesn't really help anyone. The clients don't want to pay for that time. They don't want you spending that time. They would rather focus on what can guide and grow their business. And now the technology does exist to do this. And so to quickly hit on this bookkeeping we talked about with predictive models. We now have a predictive bookkeeping model that's at 98% accuracy, dramatically outperforms every LLM out on the market in large scale benchmarks. Reconciliation. The document extraction models have now largely solved. You can automatically reconcile bank statements to the pixel bounding box on the PDF. So you just mouse over the transaction and see it on the PDF. There's a lot of work happening this year on schedule automation, on review, quality control, anomaly detection and so on. And then of course, the reporting, the analysis, the insights is just statistical analysis. And so you can use math computers to do that analysis and, and then hand the results to an LLM to summarize it and pull out the insights. And so if you tie all of those together, I would say by 12 months from now, end of year, the month end close will be 95 percent automated.
B
Yeah, so very exciting time. I think in a future town hall we're gonna bring somebody on from FASB setting the standards and the technology companies that are automating a lot of the, complying with the accounting standards and doing things like automating the close here. So a bunch of questions came in. You know what we had on our last town hall, we were talking about the modernization of the irs. The one thing that is known is a lot of the IRS code is in COBOL and fortran. So there's a lot of discussion that needs to be rewritten. And this is a huge opportunity now because it's hard to find COBOL and FORTRAN programmers and you don't necessarily need to anymore.
E
This is the opportunity they've been waiting for because these coding models are exceptional at porting code from language to language. And now they do have the opportunity to write tests for the COBOL code, have the AI port it to a more modern language, and then verify that all the tests still pass.
B
I'll just conclude by saying we have a lot of events where we're connecting with the technology community. Digital CPA Engage is coming up. You'll be there. We're three years in now to ChatGPT. It seems like the solutions are arriving. A lot of comments here. This is not something that the firms or people in business and industry you need to personally do the building of these solutions. These solutions are coming to market. So a lot more to come. And Jeff, really appreciate these insights and we're going to share in our newsletter a video of you kind of doing that. You did Digital CPA reviewing these different areas. And we'll get these resource tools out as well via the newsletter.
D
Yeah.
E
Thanks so much for having me, Eric. I will say 2026 is the year this is real for firms and I expect large scale adoption across the profession.
B
Yeah, it's moved from experimentation to kind of real practical outcomes. So thank you. Thanks, Jeff. Yeah. And here is, I just mentioned that, here's the talk that he gave at Digital. So you can see that AI Insights, we have these AI guides and our recently issued AI in Accounting reports. So this is really transforming the practice of finance and accounting and we're doing all we can to kind of help frame it for the profession. So now, Melanie, please join. And we've got an update on the irs. And then we're going to do an IRS poll.
A
We sure do. So this week IRS CEO Frank Basagnano actually said that not only is the IRS ready to meet taxpayer needs, but they have an objective to actually have higher levels of service than they have in the past. And they've also said that they are going to be able to meet large collection needs during this period too. Now all of this can happen if they rely heavily on technology. And that is part of the plan. Also, he announced at an ABA meeting that he would be restructuring the entire irs. And so what those implications mean is that he's having 16 IRS C suite level executives directly reporting to him. So he's really looking to have people try to solve taxpayer problems. And through our conversations with the irs, I've listed some of their top priorities that they have. These are immediate top priorities. No surprise. HR1 is on this list. And like I said, they're trying to resolve these problems. And part of it, too, they're discussing the standards of measurements that they've actually used in the past to measure performance rates, which they've always been an issue for me because they're not very straightforward to understand. However, some of the new performance rates measures have been critiqued because they are solely focusing on one account management line instead of the whole scope of lines that the IRS works with. So we will continue to be monitoring this closely, and as always, we need feedback from our members as they start going through the filing season. So talking about feedback, you said that we will be polling our members, so we're going to have a couple of polls here. The IRS actually reached out to us and they want feedback. Eric?
E
Yeah.
B
Just one thing, too. We're going to push the polls, but everybody knows it just takes a little bit of time for everyone to receive the polls. We ask you to answer the poll, not to give us the answer in Q and A. So sorry, Melanie.
A
No, absolutely. And so we really are looking forward to getting responses because this will be direct feedback directly to the irs. They want to be able to change and make changes. So the first poll that you'll be seeing is what is the top phone line where dropped calls occur most often? They're trying to pinpoint where this is happening, on which service line, if there's one in particular. So the options.
B
Go ahead. Go through the options, please. Yes.
A
So the options are the PPS line, the priority practitioner service line, the general 1040 line, the automated collection systems and other specialty items. And then for all the poll questions, we are going to have the option of not applicable if you don't call the irs.
B
Okay, so we have three polls. So this is the first poll. We'll leave it up for just a couple. Couple more seconds. And as you said, these are all related. They're trying to fully understand how these drop calls are occurring.
D
Right.
A
And they're trying these questions. They're just trying to pinpoint where the problem is so that they can make a change. And it's actually the first time the IRS has reached out to us asking for direct feedback from our members. So this is great.
B
Well, let's go to the second poll. So we'll push that live. Sorry. And this is going back to two.
E
Yep.
A
So the question is, when does the call get dropped? And your options are while on hold, before you're able to connect to an assistor, when speaking with a representative, or while being transferred. And Then of course, not applicable. Now, I'm hoping that if we are able to give a decent amount of responses back to the IRS and feedback, that we'll be seeing more of these types of polls coming from the IRS in order to fix other problems.
B
And once again, the poll should come to you. Sometimes there's a slight delay and it's best you Answer the poll vs answer in Q and A. So when does the call get dropped? So we'll leave that up for.
D
A.
B
Few more seconds and why don't we move to poll question number three?
A
Yeah, so the question here is, what are people calling in for when the calls get dropped? Like, what is their objective of these calls? So your options for responses are when you're asking the assistor to do certain things. Things, meaning you want them to take certain actions. The next one is asking for information to determine the cause of an issue. So you're looking for information from the IRS to have better understanding of the issue. The next option is to be directed elsewhere for assistance. So in other words, you're looking to be transferred somewhere else within the IRS system and of course, not applicable for this. And we do know there's been a lot of groups that have reached out to the IRS because calls are being dropped. And it's frustrating because there is a courtesy disconnect and oftentimes you can be on hold for a long time. You get dropped, and then you have to wait in line again. So hard times.
B
Okay, well, excellent, Melanie. We look forward. We will share these results with the town hall community and the irs.
F
Sure.
A
So let's go ahead and start talking about IRS funding, which we had a quick or mark covered real quickly. On January 11, the appropriations released the Financial Services and General Government spending bill. And within that bill, it had the IRS fiscal 2026 budget. Now, from what you can see, there was a 9% cut to the budget from prior year to this proposed or 2026 funding. Now, 2026 is more or less set right now, so I don't expect to see a lot of movement. One thing is, yes, it is a 9% overall drop, but we miss the worst that it could have been. And the worst case scenario is what you see in the middle column there. But there are things to note with the budget here. Now, we do know that Technology and Operations actually received a 23% cut. And technology and operations is lumping two units before which was the Technology and Operations supply support into one. We also are seeing that enforcement is getting an 8% cut, but most importantly, Taxpayer Services actually got a boost of 9%. So that is something good to know. Now another thing to Note is the 1.1 billion shortfall. The IRS actually has the funding to cover that shortfall from the Inflation Reduction act funding that was given to them. Having said that, there is another spending bill which is the Health and Human Services spending bill, which usually there's nothing tax in those bills. But they are proposing to claw back almost $12 billion from operations support. And a report actually came out today saying that could have an impact of 38 billion in lost revenues. So all these things are things that we are going to continue monitoring and taking a look to see how it plays into the filing season. Moving on to the next slide, mandatory electronic payments. That is a question that we get a lot from members. We're still waiting on guidance and I know we're four days out from the filing season start of it and people they have told us guidance would be dropped but unfortunately we have not seen that guidance. But I do have some updates for you based on conversations that I've had with the irs. So first the business returns, they are going to have a box where people can put in their direct deposit information. 2 We also know some exceptions. They did confirm that if you are a US Taxpayer living abroad you will be accepted and if you are are an estate or a trust, you will be accepted. Which those two areas have been areas of concern. There will be another greater list. But for now that is what we know. Another thing that we know too is that the IRS is looking at a variety of ways to fund the refunds. So we obviously know about direct deposit and the prepaid credit card but they're also looking now at Zelle Venmo PayPal. So they are looking to broaden that to make things a little bit easier. And we've also learned a little bit about the process of what we are going to see. So when you go to file the return and if you don't provide your direct deposit information, the software will give you an alert that you didn't provide this information. That doesn't mean your return can't be processed. It doesn't mean it won't be submitted. Everything will move forward but you can expect to get a notice and in that notice it will tell you how you can provide the direct deposit information for them. Now the thing that's interesting about it is if you don't provide that information at the six week mark, the IRS is going to cut a paper check of a refund to you. So again we are waiting for additional information for now we have resources listed on this slide and that last link there is for an AICPA resources and that's where we will be updating it as we get more information from the IRS. Moving on to the next topic, it is the 100% bonus depreciation. On January 14th the IRS released interim guidance and it is well received because essentially short of the effective dates, it's pretty much TCGA rules. So if you were familiar with the TCJ rules, you're pretty much right back at it again. So that is really good. We do know that there are some differences though. We do know that first people can elect to not take the first year depreciation and they can start to take it in 2026. And the other thing that's new is that they did make a new section for sound recording qualified sound recording productions and details are up here on the slide. And we also extend expect that this guidance will be consistent with proposed regulations and final regulations. So all of it good news. Moving on to the definition of a limited partner and I did see some comments coming in but real quickly, yep, on January 16th the Circuit Court, the Fifth Circuit Court actually came forward and overruled the Tax Court's holding in the Series and Solutions case regarding the definition of a limited partner for purposes of self employment tax exception. Now the what a limited partner is has been percolating through the tax courts for many years and essentially the tax courts came forward and said, you know what, there needs to be a detailed functional test which was really fact driven and burdensome to do. And the fifth Circuit Court said, yep, simply put, it is the definition of whatever the state defines it to be. So that does create certainty for those that are within the 5th Circuit Court. And we will be following those because, because those that aren't in it, we would hope to see a little bit more clarity and simplification there. Okay, moving on to an issue. The required minimum distributions rules for inherited IRAs. A lot of people think that there will be forthcoming penalty relief and there won't. Final regulations were issued on July 18, 2024. They had given penalty relief for a few years, but that's because the guidance wasn't there. Guidance is out and it is final. There will not be any penalty relief. And if you want more detailed information, there's a couple of links there that are open to town hall audience for two more weeks. So please take a look at that and let's dive into the round robin and I'm only going to hit the high up the most important ones. So when we flip to the slide, we're going to text base. Let me do something real quickly. Direct Pay update that has changed from two payments per day to five payments per day. And then the last thing that I want to get in is as we prepare for the snowstorms. As a reminder, we have a disaster relief resource center that you should look at it. Benevolent funds are in there for those impacted and I really hope everybody stays safe and warm. Back to you, Eric.
B
Mallory, that was fantastic. So we'll get some questions during Open Forum. Thank you very much. Lisa bring you up to introduce and run this next section. And Lisa, we're at a, you know, close to 15,000 almost, maybe it'll be a record town hall today, you know, that's awesome. I love it prior to the big snowstorm.
C
Cool. Let's dive in because this is an important topic that is brand new. We're going to be talking with April Walker, who's here with me in the Durham studio, and Nick Farr with Forbes Mazar as a top 10 firm about digital asset income tax return reporting. And we are going to be seeing soon a new income tax reporting form, so a 1099 DA, which means digital asset proceeds from brokerage from broker transactions. This is going to change how digital asset activity is tracked, how it's reported and reviewed by the IRS and by us if we're working with our clients or if we're working with our kids on. Oh my gosh, you did how many digital asset transactions this year. So as I mentioned, I've got April here and Nick, they're part of our digital asset task force and they've been doing some great work. So they're going to work with us on trying to explain what to expect and why we're having this new type of reporting. So, April, as we look at the next slide, let's start at the basics. What do we mean when we're talking about digital assets?
F
Yeah, that's a great question. And I'm going to refrain from reading, which I would have to do, read the definition out of the Internal Revenue Code for digital assets. But basically think you know, anything of value that exists on a, in, in a digital form. Right. That's on a blockchain or, or similar type technology. And examples are probably helpful. You think of cryptocurrencies such as Bitcoin or Ethereum, NFTs you probably heard about. There's been a lot of talk, we talked about it earlier today on Town Hall Stable, Stable Coins. So that's that's kind of what we're talking about here.
C
Okay, that's a good level set. So why is this new digital asset reporting form a big deal to our audience?
F
Yeah, because, I mean, there's been digital asset transactions for quite a few years that have been, you know, needing to be reported on tax returns, of course, but this is the first time that custodial brokers in the United States are required to, to issue standardized information reporting. So those are will be coming. And again, we're hoping, you know, you might be out there thinking, oh, okay, this is going to solve all my problems. This is going to solve all my problems where I didn't understand what my clients were doing or my businesses were doing on digital assets. And I don't like to be the bearer of bad news, but I don't think that is going to be the case. Right. There's still confusion. I still hear it a lot about what's going to be reporting, what's not going to be reported yet, and you know, kind of where that is. And Nick's going to dig more into that detail.
C
So, Nick, that's a great call out for you. As we look at the next slide, what should we expect to be reported? To not be reported. And talk to us about what 2025 is going to look like.
G
Yeah, of course. And first of all, thank you all for having me on. It's exciting to be here. So at a very high level, custodial brokers, and I use that word custodial brokers very intentionally and April did as well, are going to be reporting digital asset transactions on 1099s for the 2025 tax year. Now, you might be wondering, what is the custodial broker? What does that mean? So let me use an example. When you put money, or when you have money, let's say you get paid in cash, just for example, you have the option to keep that cash in your wallet or your purse or under your mattress, if that's how you like to store your cash. Or you have the option to go to the bank and deposit it with a bank. And in that case, when you deposit it with the bank, the bank is the custodian. Okay. So similar concepts for crypto. When you own crypto or digital assets, you can store it in your own wallet where you have full access and control over it, or you can store custodian. Custodian could be a centralized exchange. There's several in the US that are very common and well known. And so in this case, only custodial Brokers will be issuing these 1099s for these digital asset transactions. And so because of that, there are several different types of transactions that actually won't be reported on a 1099 for 2025 and even going forward in future years. And so anything that occurs through a self hosted wallet, which is not with a custodial broker, won't be reported on a 1099. There's also certain transactions that the IRS has specifically identified in Notice 2024-57. There's about six different transactions in there that custodial brokers do not have to report on these 1099s as well. So while this is the first time that these transactions will be reported on 1099, it won't be all encompassing.
C
And that's why that reminder at the bottom of the slide, taxable digital transactions must be reported with or without a 1099da. So I always love that, that we call that out. So, so when we get the 1099DAS, is that going to be a complete accounting, a complete transactional record that taxpayer tax practitioners can just use and prepare the returns?
G
Yeah, unfortunately, the short answer is no in most cases. April, as you mentioned earlier, the intent of this new 1099 reporting regime is to raise revenue and reduce non compliance there. It very well could be many clients out there that haven't reported these transactions in the past that may now start to receive these 1099s. And they may come to you as a practitioner and say, oh, you know, Here's a new 1099 that I received and you may start digging in and asking questions and they may have had transactions that occurred even before 2025. So unfortunately, you can't just rely on the 1099s because it won't be all encompassing of all of the transactions. And in addition to that, there are certain information that may not be reported on them as well. For example, cost basis does not have to be reported in 2025. And so there may be 1099 that are issued where there is no cost basis or missing cost basis.
C
The dreaded missing cost basis. I'm like, I'm going to break out in a sweat. And it's been 20 years since I've had to deal with that. That's always terrifying. So, April, what does this mean in practice? What should our practitioners be doing right now to get ahead of this?
F
Yeah, hopefully they've been preparing for this. We've been talking about it for some time. But two things that I want to Highlight that I think are really important and one we've already highlighted. But I sometimes, you know, when you hear it again, the second time, it hits a little better. But reporting obligations haven't changed. Right there, there has been transactions before, there's transactions now. You can't just put 1099 DA on the tax return, call it a day and move on. You need to make sure you're still asking the questions. You're, you're trying to get information. And you know, there, there is possibly some delay in the 1099 DAs that you might receive. And so that's still, you know, that's not going to change the fact that it still needs to be reported on the tax return. So that's number one. Number two is that you need to be prepared as a business or as a practitioner to deal with these challenges, these data collection challenges. Nick mentioned it like cost basis is not required to be reported. If your clients don't have their cost basis information together and ready and document, you know, substantiated their cost basis and substantiated the cost basis is zero. I mean, that's not a great answer. That's not a great answer for anyone, but that is the answer if you can't substantiate. So helping your clients, whether it's with software or getting them hooked up with an expert if you're not able to deal with it, is really what we're recommending.
C
So Nick, as we move into talking about some of the resources that we have, let's hit one more topic really quickly. If you've got a client and boy, again, 20 years, but I'm still going to break out in a sweat where they come forward and say, so here's this and I didn't tell you about it for the last three years. Now what. So what would you suggest that our practitioners do in those circumstances?
G
Yeah, it's a great opportunity to go back and look at the options of amending returns and rebuilding that cost basis. Because as April just mentioned, if there is no record to rely on, the IRS in the worst case will assume zero cost basis. And so as alluded to, there are third party softwares out there that are available that can help automate some of this process where you connect the clients, all of the clients, wallets and exchanges, and will help aggregate the data on your behalf. Now, I would encourage you to make sure, do research on those, make sure they're SAW compliant, make sure the calculations are correct and accurate and complete. But that is a good place to start.
C
Okay. I think that's great advice. And if you've been part of the town hall community for a while, you know that I always like to offer hope when we have some of these challenges that we're presented with. And we have an amazing resource that Nick, April, and other members of our task force have been working on. So please check out the Digital Asset Tax Framework practice aid.
F
That's a mouthful.
C
And they've made the first couple of chapters available to everyone so that you can see it, and then you'll. If you're a tax section member, then you have access to the full content there. We have a ton of other resources available that you'll find a link to on the slide here. Eric, how are the questions looking?
B
It's a good day for questions. So on the digital, you know, the 1099, the 1099 DAs, like, when will a bunch of these. When will they be distributed? You know, when will. When will they be coming out? So all different versions of that question.
C
That seemed like such an easy question, you guys.
F
Right, I'm gonna let Nick take that one.
G
Yeah, yeah, yeah, I can take this one. So similar to other 1099 Bs that brokers issue, the general deadline is February 15th. Now, that falls on a weekend this year, so it'll be February 17th is when the actual due date is, so you can expect your clients to start receiving them around there. Now, since this is the first year that these 1099 DAs will come out, there is temporary relief for the first year where if something goes wrong and these brokers act in good faith, they can apply for up to an extra year to issue these 1099s. So that's where April alluded to, even if your client extends their tax return to maybe October 15th, they could receive one of these 1099s past that filing due date.
B
All right, all right, thank you. Well, why don't we move into Open Forum? So advance the slide here. Bring Mark and Melanie back up. So at least as you're looking for a question or. Melanie, I don't know if you've seen a question that you want to take. I'll just say the leading category. We break these all into categories. We got over 100 AI model questions that have come in. And just a general theme here is, how do I keep up? Again, some questions based on. On firm size. What you can be confident about is you saw Jeff Seibert from Digits. There's a whole host of technology companies. Some of them are new. Some of them have been in place for a long time and their names that, you know, they're all over this. So they're all over bringing these new capabilities to market. I think looking at some of these frameworks that we're putting out in these toolkits on how to think through selecting these different partners, vendors, evaluating them, that's important. Understanding how they're providing this service, Are they using generative model like ChatGPT? Is it predictive modeling? Is it document extraction? So I think just building more of an understanding of the category and going to events like Engage or staying connected with us on the town hall. And I know there's lots of questions related to recommendations, so that's something we can take back and think further about. So, Lisa, is there a question that you want to grab?
C
I just continue to see a lot of questions about electronic payments. And Melanie, I know you went through that litany of both IRS and AI system CPA resources. So I'm just going to refer you back to there and know that Melanie and the team are working really hard with the IRS to continue to talk through some of the challenges that you're bringing forward. And I know she'll keep you updated on that.
A
And I just want to add that we do not have formal guidance from the IRS for payments to them. So until we get that guidance, paper checks are allowed.
B
Yeah. And that's one top question. When will IRS add trust in estates through the direct payment system that came in?
A
Yeah, that's a whole other issue. Modernizing the irs.
B
Okay, well, Lisa, why don't you and I move to the resources section. Mark, Nick, Melanie, thank you. April as well.
C
I'm going to start us off as we head into the closing with a reminder to go have some fun, y'.
E
All.
C
Work is work is work, but you can, you can work in some, some busy season fun as well. So PCPS provides one of our most favorite, most popular resources every year. And these are some ideas for you to have fun with your colleagues. Don't have to just be in a public accounting firm. It's a busy time of year for a lot of us. So look at these ideas and have some fun.
B
Here's some more resources, some of the benchmarking reports that we've recently released. It's always good to see how other firms are approaching audit transformation their CAS area. And also there was a number of questions that came up related to tax research. So here's some tools that you can leverage. The next Town hall coming up on February 5th at 3pm We've got a great lineup here Lisa, I'll let you comment on one of the topics. I'll comment on Jeff Weiner coming on. Jeff was the founder basically and former CEO of Markham. He's going to give us some reflections and insights on how to drive firm growth.
C
And Lexi Weber is our Emerging Partner Senior Manager. She's going to be chatting with Lauren Baptiste about our new PCPs and emerging professionals Burnout Prevention Toolkit. There's some great topics to cover in that conversation, and we're excited to be bringing that to you in the Town hall.
B
And we always have our Town hall resources@cpa.com townhall so great being with all of you today. Stay safe, stay warm and look forward to being back with you sometime soon.
A
Thank you for your participation. You can also subscribe to the AICPA Town hall series on your favorite podcast platform, as well as watch archives on YouTube and find resources@cpa.com Townhall Tune in for live broadcasts Thursdays at 3pm Eastern Time.
D
This podcast is designed to provide illustrative information with respect to the subject matter covered and does not represent an official opinion or position of the AICPA or AICPA.org it is provided with the understanding that The AICPA and AICPA.org are not engaged in offering legal, accounting or other professional service. If such advice or expert assistance is required, the services of a competent professional person should be sought. The AICPA and AICPA.ORG make no representations, warranties or guarantees as to, and assume no responsibility for the content or application of the material contained herein, and especially disclaim all liability for any damages arising out of the use of, reference to, or reliance on such material.
Episode Title: Digital asset reporting and technical updates ahead of busy season
Host(s): Erik Asgeirsson, Lisa Simpson, Mark Peterson
Featured Guests: Jeff Seibert (CEO, Digits), Melanie Lauritson (AICPA IRS Liaison), April Walker (AICPA), Nick Farr (Forvis Mazars)
Main Theme:
This episode delivers a comprehensive update for accounting professionals on three major fronts:
Presenter: Mark Peterson
Timestamps: 02:03–13:56
Geopolitics’ Expanding Influence:
Midterm Election Dynamics:
Tariff and Trade Uncertainty:
Government Funding and Shutdown Prospects:
Digital Asset Legislation:
Federal Fiscal Responsibility:
Student Loan Cap Issues:
“If there’s ever a time not to just be chasing every headline, it’s right now.” – Erik Asgeirsson (04:58)
Speakers: Erik Asgeirsson, Jeff Seibert
Timestamps: 14:07–31:35
Jeff Seibert explains a taxonomy of five AI model categories for accountants:
Generative AI (LLMs, e.g., ChatGPT):
Predictive Models:
Document Extraction Models:
Agent-based/Agentic AI:
‘Not Actually AI’:
“2026 is the year this is real for firms and I expect large scale adoption across the profession.” – Jeff Seibert (31:28)
Presenter: Melanie Lauritson
Timestamps: 32:17–44:17
“This is the first time the IRS has reached out to us asking for direct feedback from our members.” – Melanie Lauritson (35:24)
Speakers: Lisa Simpson, April Walker, Nick Farr
Timestamps: 44:39–56:08
Timestamps: 56:08–58:45
For actionable resources and continued updates, stay connected via CPA.com/townhall.