The Anthropic IPO?

0:04Good morning, good evening, wherever you are. Welcome back to the Enterprise AI Show. I'm your September, 2026, sort of where they stand as they start to prepare for potential IPO, whether that is sometime this year or it gets moved into next year, whatever happens, sort of broke down their financials, sort of broke down their strengths, their weaknesses, you know, where they were revenue wise, where they were in terms of financial obligations, funding, all those sort of things. And thought we would do the same sort of thing for Anthropic, seeing as they are, you know, the two of them are really kind of considered probably number one and Number two, in terms of Frontier Labs, sort of front and center in terms of a lot of the revenue, whether it's direct revenue, circular revenue, circular funding, whatever you want to look at it in terms of the AI industry.

1:05Obviously, there's other players like NVIDIA and Google and others, but these two are sort of front and center. And we thought we would kind of dig into the Anthropic case right after the break.

1:18Today's show is sponsored by Nasuni. There's a growing gap in AI right now between what's possible in theory and what successfully works at scale inside an enterprise. The difference comes down to unstructured file data. Many AI initiatives struggle because the file data they depend on is scattered, unstructured, and disconnected from where and how work actually happens. Nasuni changes that. It brings your unstructured file data into a single, secure foundation so AI, both generative and agentic, can access it with the context, governance, and performance it needs in production. Bring AI to where your unstructured data lives.

1:52See what it takes to activate your data for AI and request a demo at nasuni.com slash AI.

2:01What does the outside world already know about your company? Cyber criminals could be seeing leaked credentials, compromised session cookies, exposed infrastructure, and even impersonations of your brand and executives. Nord Layer Intelligence by Nordstellar gives security teams visibility into those threats across the deep and dark web, data breaches, attack surfaces, and brand impersonation all in one platform. Find out what attackers know before they can use it. Visit nordlayer.com slash intelligence slash enterprise AI and use code enterprise10 for 10 % off your Nordstellar plan. And we're back.

2:47And as I mentioned in the show on Sunday, nothing in this show should be considered financial advice. This is not intended to be financial advice, forward-looking information. This is really just us looking at information that is out in the public or relatively public. Like, you know, sometimes these are numbers that are derived from leaked information and so forth. But a lot of folks trying to just get a handle on kind of what does the picture look like for Anthropic in terms of as they start to get close to an IPO, you start to look at things like, you know, not just is their technology cool?

3:18Do they have a vision for the future? Are they attracting great talent? How fast are they able to build new models? But you start looking at the financials of the business. You look at the go-to-market of the business. you look at a lot of things because as we move from sort of fully VC funded, round to funding, start a startup phase into something that's going to go into public phase, you start to have a whole lot of different kind of scrutiny upon you. And some of that scrutiny is just from the timing perspective of moving into a public company, quarterly earnings, quarterly demand, starting to have outside investors who have very large stakes in what you're doing, whether those are hedge funds or they're institutional investors, or at some point they could become sort of raiders and hostile investors and so forth.

4:09So the move from being a private company to a public company does just come for any company, not just for Anthrop. It comes with a bunch of scrutiny and it comes with kind of a push for a faster timeline than you typically had when you're a private company. And you sort of controlled your burn rate. You sort of controlled your roadmap. You controlled how fast you were putting these out in the future. And obviously, the marketplace plays a role in that as well. You can't just decide that you're going to do things whenever you feel like it. There are competitive threats, competitive pressures. Customers are going to make decisions that fit what they want as opposed to what you want.

4:48But I thought it'd be worth digging into where Anthropic is these days, especially as we are in end of September, 2026. There's talk that they may try and go IPO here in the November timeframe. And again, these are all just speculative things. Please don't take them as financial advice. But if we dig into some of their numbers, we dig into kind of where they are today. If we look at the numbers that have been put out there, and again, keep in mind, these are numbers that are kind of coming from a lot of different places. In some cases, they're perceived to be leaked numbers that might've got picked up by the press.

5:22In some cases, these are numbers that are kind of being relayed from investors that are getting pitched to potentially be larger investors in Anthropic or being pitched to banks to be the listing agent for Anthropic as they try and go IPO on one of the major markets, whether it's New York Stock Exchange or NASDAQ or whatever it might be. So just looking at a few of those numbers, first thing we want to look at is sort of top line revenue. The numbers that are being put out there, if we go back to 2024, estimated revenue was somewhere in the$380 to$400 million range. We go forward a year to 2025, estimated revenue is perceived to be up about$4.6 billion.

6:07So roughly a 10x increase in a year, huge, huge numbers in terms of growth rates. And then as we're looking at 2026, they put out a number or they're floating around a number of about$11.5 billion for Q2. They're trying to say their run rate is somewhere around$60 billion. Usually what you would do is you would take the latest quarter and multiply it by four. So let's put it somewhere between$45 billion and let's say$60 billion because they can sort of see what their growth looks like over the next couple of quarters that we can't see. So if they went from$4.5 billion in 2025 to potentially$60 billion in 2026, or even if it's$45 billion, that's another 10x sort of jump.

6:55So they went from, again, putting these back in context, 2024, let's say$400 million, 2025,$4.6 billion, 2026, somewhere between$45 billion and$50 or$60 billion. So at least a 10x jump the last two years in terms of revenue. So that's going to be in the top one, two, three fastest growing companies of all time. Now, on the flip side, 2024 operating loss, roughly$3 billion. So made$400 million, lost$3 billion. 2025 made$4.6 billion in revenue, lost a little over$8 billion in revenue. So that's about$11 billion over two years against, you know, about$5 billion in revenue. And then going forward, you know, they didn't put out numbers for 2026.

7:50You know, you have to assume if they went 10x, you know, so if you look at 24 to 25 revenue numbers, they went 10x. They more than doubled their operating loss. If they did that again, if you assume, you know, another 10x, but double their operating loss, you know, they're looking at an operating loss probably somewhere in the, in the range of 15 to$20 billion, uh, for, for 2026. And that's on$60 billion of revenue. So, you know, it's a business that is very expensive. Um, you know, at least at this point in time, um, you know, it's a different business than we are used to having measured from the sort of.com, uh, and web era, uh, where everything was kind of driven by, um, you know, if you're a web company, for the most part, if you were an enterprise company, you know, it was enterprise adoption, you were a SaaS business and so forth.

8:40Your costs were, you know, still sort of very similar. Your costs were in sales and marketing, you know, and so the nice thing about that business is people really started to understand it. They understood that at some point in time, once you started to have customers, you had sticky customers, they were beginning to stay with your SaaS service. For example, you could cut way back on sales and marketing. You didn't have to keep aggressively going after that same customer to grow what they were doing. You would use sales and marketing dollars to go after new customers, but those recurring customers became very, very profitable, very sticky.

9:12This business is different. We don't yet know how sticky and what the growth rate is for any given customer. We can sort of project it based on some of the numbers that we're seeing here in terms of the 10x revenue growth, but the cost of this business is very different in that you have massive cost of having to make upfront commitments against GPUs, which are kind of a...

10:01Launch it at a target of the highest bidders choosing. Get the news from the end of the world. Follow the clues. Unravel the mystery as civilization collapses. 2029, The War for Our World. A science fiction news series from the future. Premiering October 30th. Follow wherever you get your podcasts.

10:23You know, in-demand resource or a finite resource right now, you know, shrinking resource. So we're seeing them make huge commitments into data center providers, whether it's AWS or Google or now SpaceX going to the future. So you've got sort of raw material level things. You've got the cost of extremely expensive data engineering teams, data science teams and so forth. And then you've just got the cost of training, which some of that is built into the cost of GPUs and the forward sort of expectation and commitments of GPUs. So it's a business that is very expensive to be able to turn out the raw capabilities that you're delivering.

11:08Whereas in the past, software was very much expensive on the front end. You build the software, you build capabilities. But once you reach sort of a steady state point, they were highly, highly profitable, 85%, 90%, 95 % sort of profit margins beyond that. So right now we're looking at a business that is growing revenues very, very rapidly, which is a very, very positive thing. They are growing operating losses, maybe not at the same rate, but they are still growing operating losses. operating losses haven't necessarily shrunk, although they are potentially coming down as a percentage of revenues.

11:47It is an expensive business to be in, both in terms of CapEx, whether you are building your own data centers. And for the most part, Anthropic has not made commitments to build their own data centers. They are still primarily renting GPUs, like I said, from AWS, from Azure, from AWS, from Google, and now from SpaceX. So a little bit different cost structure than we see from some of the other providers like Oracle and OpenAI who are making huge commitments to build and own their own data centers. So different financial commitment, different timeline of where your capital is tied up in. Anthropic seems to be more of a renter.

12:30that is advantageous to them to a certain extent that they are not making the huge commitment around the data centers and some of the slowdown around data centers. It also puts them at a little bit of a risky situation in that the people that they are renting from who they are becoming massive, massive customers are, are also some of their largest competitors, whether it's AWS or Google, or to a certain extent, like Grok with SpaceX and so forth. So, you know, something to keep in mind going forward. Just looking through some of their numbers, you're going to see some things. They had a net loss in 2025 of$42 billion.

13:10A lot of that was driven around what they call non-cash accounting. A lot of it has to do with being able to pay people equity, putting equity into the business or having expense of equity for people that are part of the business. And so So we're going to see that number quite a bit from OpenAI, from Anthropic, from anybody who's trying to go IPO is huge, huge amounts of equity and the value of that equity that they're having to account for as they get closer and closer or those things are being sold in the secondary market. So that's a number that you kind of have to keep in mind is a little more complicated to explain versus your normal sort of like revenues, cost of goods sold, cost of sales.

13:53But that's a piece that's there as well. So it kind of gives you a sense of where that is in terms of their future commitments. Right now, they're on the books for$518 billion of long term infrastructure obligations. I think that's through 2030 or 2032. So over the next, call it five years, five, six years, they've got about a, you know, roughly on an annual basis, a little more than$100 billion of forward commitments that they're going to have to make. So as we think about, you know, the profitability of their business, you know, currently not a profitable business. I know they're trying to play some games in terms of saying, well, we're profitable except for, you know, all these costs.

14:35But net net of it is right now still not a profitable business. And they're going to have to get to a point where they are, you know, net profitable of at least$518 billion, you know, to deal with these long term obligations. they might be able to be able to get out of those contracts or reduce them over time. Some of those costs might come down, who knows. But, you know, huge, you know, huge amount of long-term commitments that need to be, that come out and need to come out of profitability, profitability after tax, you know, and right now still not a profitable business. So if we look at funding rounds, they have raised over$130 billion across, they're now at a series H.

15:17So what is that? About seven or eight rounds. A little more than that, depending on how you do letters. But anyways, about about one hundred and thirty billion dollars. Like I said, you know, loss in the last couple of years is, you know, somewhere on the order of let's call it thirty billion dollars loss. loss. So they do burn through money pretty good. Do have some available cash available to them. But a lot of this seems to be sort of like all tied up at this point. There's not a whole lot of flexibility. It's a little bit of flexibility they'll have as they go IPO to be able to use their equity for things like acquisitions and things along those lines.

15:56So probably most likely smaller acquisitions that are going to help them move further up the stack as they look at what they're doing. And then they've got about$71 billion in off balance sheet commitments. So things that are sometimes called special purpose vehicles, mostly through Google and Broadcom. These are things that they have commitments. They are backed by other financial institutions like Apollo Global, Blackstone, some others, to be able to get them access to other things in which you would expect to spend your capital on. But this is sort of off the books, right? So you can, you can sort of think of adding that 71 billion to the other 518 billion they have commitments to.

16:38So that, that, that future commitment is roughly$600 billion again. So as you're trying to figure out like, what does this company look like? You know, the market for generative AI needs to get probably significantly bigger than it is today in terms of them being able to get to a point where they are profitable and can keep up with the commitments that they need to make just to expand the growth they are sort of pushing out into the marketplace. As I look at sort of the breakdown of the company, sort of strengths and weaknesses as we did with OpenAI, a couple of things come to mind, a couple of things jump out.

17:16I think in terms of strengths, they do have extremely good brand awareness with software developers. I think if you talk to software developers, Claude is Claude and Claude Code and sort of the tools that are now around Claude are pretty ubiquitous with software developers. Whatever that number is these days, whether it's 10 million software developers or 25 million software developers, whatever the number might be, we're seeing it expanding out. They do have a very strong brand in the enterprise. So white collar workers, software engineers, researchers, you know, they were very, very quick to basically identify that, you know, software development is their core, you know, kind of core bread and butter research.

18:03And then they're trying to expand that out into obviously a number of other white collar works. We've seen them build tools for financial analysts, for lawyers, for, you know, a lot of other things for marketing, for other white collar roles. So a lot of the, you know, anthropic, you know, future needs to be probably very less about sort of augmenting what people are doing and very much about, you know, trying to replace what people do at some price point that is significantly lower but more productive than what people are, you know, what companies are paying for individual humans at this point in time.

18:45Also, huge amounts of strength around depth and experience of their frontier labs. You know, again, you know, considered in the top two, maybe one, maybe three, depending on if you include Google in there. But, you know, very much considered world class in terms of their engineering skills. They do seem to be able to attract very good talent. You know, Luz have lost some people from time to time, but people tend to sort of gravitate back there. Um, in terms of weaknesses, you know, I think the biggest, uh, the biggest concern that people probably have at this point in time is, is, is a couple of things.

19:16Number one is, um, you know, Dario Modi, their CEO, uh, frequently tends to, um, scare the shit out of the world with his proclamations about doom. I'm sorry about the language there. If there's kids in your car. Um, yeah. So, you know, he's, uh, it's an interesting stance in terms of, um, you know, he, he, he writes, he tends to write these sort of long, you know, big idea sort of things. But then when he gets around to talking, you know, he tends to come back all the time to this idea of like, well, you know, there's a good chance that this destroys the world, that this causes, you know, all sorts of problems for things.

19:54And so, you know, they're probably, as we go back to, you know, the basic idea that your stock is, you know, a combination of what you actually do in terms of revenues and profit and loss, but also, you know, multiply that by the story that you tell. And it's been a while since Anthropic was in the news for talking about sort of big picture things, talking about, you know, their technology was able to help, you know, cure a major disease, which was, you know, kind of the talk three or four years ago when then they were talking about, you know, AGI and getting to AGI quickly and so forth. So, you know, they've got some work to do in terms of, you know, what they're perceiving out into the world as to what they could be, especially as we get beyond just tools that help knowledge workers, you know, augment what they do on a day to day basis.

20:38I don't think they're necessarily going to, you know, make up going to be a two trillion dollar company, probably trying to be a five trillion dollar, ten trillion dollar company like NVIDIA, you know, doing nothing but selling 20 and 200 dollar subscriptions to a bunch of white collar workers. They're going to have to figure out, you know, how to turn it into a much more productivity, creativity, creating a new idea, creating machine than it does today. You know, the other, the other big sort of weakness is, I think there is a perception that they, at least with the current administration and a set of the Silicon Valley sort of elite, if you will, that they're sort of on the other side of that, that, that fence, if you will, right.

21:23There seems to be, concern with Anthropic within the US government, how much of that is actual concern versus just kind of like showboating or whether that is pandering by OpenAI or by others who are trying to make sure that one company doesn't get out ahead of the rest of them. We've seen some of Silicon Valley and different big tech companies sort of feel like they are trying to pile up, to gang up against Anthropic. So, you know, it does feel like there's a certain amount of sentiment within the U.S. government in certain circles of the U.S. government. There does seem to be a sentiment within the venture community, as well as, you know, some of the large companies that compete against Anthropic, you know, that Anthropic is sort of competitor number one, in some cases, public enemy number one.

22:15And, you know, what that means in the future, it's unknown, right? It's an uncertainty going into an IPO as to, you know, will we see more things happen in which, you know, other competitors, others on a different, you know, perspective as to where, you know, AI should be and so forth, create situations similar to when Mythos became, you know, on the list of things that couldn't be exported outside the United States or couldn't be used by certain industries and so forth. You know, is that something that, that we can expect to see happen more frequently in which, you know, the equivalent of an outage of their service.

22:55So I think those are the two big risks that they have beyond just the financial risk of the numbers that we, that we talked about in terms of revenue today versus, you know, kind of cost of delivering that service versus their, you know, their committed spend over the next three to five, you know, five to six, seven years. And then the rest of it, I think there's, you know, there's, there's sort of just, you know, technology concerns, where does the technology go? You know, what does their ecosystem look like? Do they become somebody who is a very ecosystem friendly company like Microsoft was back in the day where everybody was sort of building on top of their platform?

23:32Where do they become more of a, you know, know, ecosystem eating type of company, similar to what Amazon did as they began to get a little bit bigger and they were, you know, AWS was sort of eating all of their, you know, all of their competitors or all of their smaller startups that have built upon them that could become a feature of AWS. So I think, you know, that's something that we'll have to watch over time. Obviously, we've seen the framework with, with Cloud Code and some other things, Cloud Work and, you know, what they're trying to do. I think they probably keep pushing on that. I do feel like a company who wants to be very vertically integrated, but it'll be interesting to see how that evolves as well.

24:10Their ecosystem play, where they become more of a horizontal player, allowing companies to build up on top of them versus when do they start competing against them when those things become successful. So anyways, hopefully this was useful just to kind of give a broader picture of, you know, one of the biggest companies in the world, one of the fastest growing companies in the world, getting ready to go IPO. Their IPO will obviously have, you know, major impact on the global economy. You may not necessarily invest in it directly, but there's a good chance that index funds will have to invest in it as it grows big enough, becomes part of S &P 500 or whatever other sort of indexes that are out there.

24:52So just wanted to give you a bigger picture of things. Again, not financial advice. Please do not take this as financial advice. We hold no inside information that we passed along or that is part of our analysis, just strictly using public information that's out there. So anyways, with that, I'm going to wrap it up. Hopefully these two sort of back-to-back, OpenAI and Anthropic, what could be their IPO? What are the things to take a look at to try and build a picture of what they look like is helpful to you, whether or not you are a user of their tools or whether you are trying to determine if they should be a partner of yours in the future, or you're just trying to understand the overall market and what the status of some of the biggest companies look like right now and what that might look like a year from now, two years from now.

25:38depending on how the market ebbs and flows. So with that, we'll wrap it up. We'll talk to you next week. Thanks for listening. Check us out at theenterpriseaishow.com for past shows, newsletters, and all things enterprise AI.