Anthropic: The best AI stock you can't buy yet?

AJ Bell Money & Markets

The episode explores the potential IPO of AI giant Anthropic, analyzing its massive revenue growth and the existential risks associated with

Key takeaways

• Anthropic's revenue has seen a massive ramp-up, with reports suggesting an increase from $9 billion to a $65 billion ARR.
• The potential Anthropic IPO faces scrutiny over high infrastructure costs, estimated at $518 billion for cloud computing.

Main topics

• Anthropic IPO and AI competition
• UK Housing and Help to Buy reforms

Notable quotes

The technology may pose existential risks to humanity.

Conclusion

The episode demonstrates how both global technological shifts and loca

Transcript preview

Speaker 4 (0:09) and welcome to this week's episode of the AJ Bell Money and Markets podcast. I'm Charlene Young and joining me this week are Sarah Coles and Tom Sieber. There's a definite political flavour to this week's show with Labour conference announcements making waves on the stock market and raising some pretty big questions for people's personal finances. Speaker 3 (0:28) Yeah, so we're Speaker 4 (0:28) going Speaker 3 (0:28) to look at plans for social care reform which have come out of conference and crucially whether the sort of headline grabbing announcement on the state pensions triple lock is going to be enough to pay for Speaker 4 (0:37) it. And with Stoptober and Sober October about to get underway, Sarah's been crunching a completely different set of numbers. We'll review how giving up cigarettes and alcohol and investing that money instead could potentially add over a quarter of a million pounds to your pension over 20 years. There is always a pensions angle. Yes, there always is. But first, Tom's actually here to Speaker 3 (1:00) take us through what's been happening on the markets, including why house builders have been enjoying sort of... burn and bounce. Of course, utility stocks have been heading in the opposite direction. And with artificial intelligence back in the spotlight, Tom's going to be looking at the latest developments around Anthropic, including its prospectus, before Dan Coatsworth speaks to Bloomberg Intelligence about some of the big questions prospective investors are asking about the AI giant. Speaker 4 (1:22) Yeah, so hi, Tom. Lots to get through this week. But As Sarah says, let's start with AI because it feels like we can't go very long at the moment, not even maybe one or two podcast episodes without it cropping up somewhere, right? Speaker 5 (1:34) Yeah, absolutely. Thanks for having me on the podcast. You're right to say that once again, AI has been kind of dominating the agenda to an extent this week. We had some kind of more stories along the lines of the kind of concerns about the safety of AI. So OpenAI actually kind of... scrapped or at least hit pause on its next model over safety issues this week but as you say we've also had quite a significant development in that there's been some reporting at least around anthropics prospectus so this is the document a company puts together explaining how its business model works and outlining all the risks before it joins the stock market and anthropics planning to do that later on this year um so they are just reports but we have had some insights into to kind of potentially what what anthropics prospectus looks like uh Speaker 4 (2:26) it was the only thing that sort of stood out in those reports to you at all yeah Speaker 5 (2:29) i mean you know we we talked about the safety concerns there and i think that is the big elephant in the room really isn't it that this risk that kind of ai systems become too powerful for their own good and that they're developing at such a pace um and i think one thing that undoubtedly stood out in the reporting around the prospectus is that One of the risks mentioned is that the technology may pose existential risks to humanity. So it's hard to think of any bigger risks than that, really. I don't know if risk even really does it justice. But there were, you know, trying to sort of take both sides of this, there were reports as well that the prospectus will Speaker 6 (3:12) kind of position AI as likely to have a greater Speaker 5 (3:15) impact. on the global kind of economy and global economic activity than industrialization, electricity, or the internet. Speaker 6 (3:23) I Speaker 5 (3:23) mean, I guess you might say, they would say that, wouldn't they? Yeah, yeah. It is nonetheless, it kind of, it puts into sort of some respect to, you know, just how significant this could be. Speaker 4 (3:33) Yeah, I mean, some big food for thought there. And I think, you know, we've seen eye-watering valuations when it comes to these types. companies. Do you think investors might become a little bit more, I don't know, discerning about what they're prepared to pay to participate in these IPOs? Speaker 5 (3:49) Yeah, possibly. I think we've seen some evidence of that. And there is undoubtedly, you know, one Speaker 1 (3:56) of the big kind of worries or bits of nervousness for the Speaker 5 (3:59) market around AI is obviously how much is being spent on it and whether or not that will actually generate a proper return. Speaker 5 (4:08) And Anthropic's no exception. So the reporting around the prospectus suggests it's going to spend $518 billion on kind of cloud computing and infrastructure necessary to sort of develop its AI tools in the coming years. How Speaker 4 (4:23) does that fit into that kind of broader battle between like the major players in terms of like amounts they're planning to spend? Speaker 5 (4:29) It's a good question. It's the right question, I think. So Anthropic... is seeing very, very significant revenue growth. Certainly, with a backward-looking, with a rear-view mirror, it did. In 2025, revenue was up 12-fold. Even for an AI-related company, that is huge, huge growth. There aren't that many companies around offering that level of growth. One thing that's significant about Anthropic and maybe makes it stand out a little bit compared to some of its peers is that... It's mainly focused on kind of selling its tools to companies. And that could mean that its revenue is a bit more predictable and kind of repeatable and reliable than some of its rivals because companies will typically allow some time for kind of the technology. And obviously the main Speaker 2 (5:21) technology Speaker 5 (5:22) that Anthropic is known for is its Claude's chatbot. But, you know, it has a number of kind of AI tools. They're going to be prepared to give that time to sort of take effect and, you know, to allow them to see the benefits and how it's kind of transforming a company's processes. Whereas an individual might be a bit more ad hoc, a bit more kind of, you know, they'll use AI when they feel like it, but they're not going to be consistent and they might not be prepared to pay as much to use it. The only thing kind of on that, you know, that's obviously a kind of tick for Anthropic, but... Nearly 25 % of last year's revenue reportedly came from two customers. And that's obviously a risk, you know, because if one of those decided to go elsewhere, that would have a really significant impact on Anthropix. So I think as it joins the stock market, investors will kind of be hoping that it can broaden its client base out. Speaker 4 (6:16) Yeah, I think there's some interesting things I've sort of seen and read about businesses, excuse me, that have maybe already embedded it into their processes. And then if they're sort of tweaking the licenses, like you say, moving from like an individual to being in a firm and how they charge for that. That's an interesting point as well. But yeah, thank you. Well, that has teed things up very nicely for this week's interview. And we've given it top billing in the episode. You know, as we've mentioned, AI giant Anthropiq is expected to join the stock market very soon now, meaning investors are starting to look deeper at the company and, as we've said, work out if the shares might be worth buying. With that in mind, Dan recently caught up with Mandeep Singh, who's Global Head of Technology Research at Bloomberg Intelligence, to talk about some of the questions that these prospective investors are asking. Let's hear what they had to say. Speaker 6 (7:09) I'm very pleased to be joined by Mandeep Singh, who is Global Head of Technology Research at Bloomberg Intelligence. Mandeep, thank you very much for joining us. Hope you're well. Speaker 1 (7:18) Sure. Yeah, great to be on with you. Speaker 6 (7:21) Well, I want to talk to you all about Anthropic, potential stock market debut that we could expect potentially in the sort of the coming weeks, months. We're not quite sure exactly when it's going to happen, but there's lots of rumors. So maybe the first thing to start is... I think people might be aware of the name Anthropic. It runs the Claude Artificial Intelligence System. But what exactly does this company do? Speaker 1 (7:46) Yeah, they are one of the frontier large language model companies that has probably the best model when it comes to enterprise use cases. And we have talked about coding agents a lot this year and what sort of an impact coding agents has had on developer productivity, software engineering. And that model is the cloud. model, which is Anthropic's product. And these guys have been ahead in terms of the competition, whether it's OpenAI or Alphabet Gemini and Meta. And it feels like they have maintained that lead, even though ChatGPT was the first one in that LLM race, Anthropic came from behind. But over the past 12 months, they have maintained that lead. And now they have reached an ARR of $65 billion. They ended last year at around $9 billion. So it's almost a sevenfold increase in their annual recurring revenue to about $65 billion, which is phenomenal for any company. We've covered a lot of secular themes and tech disruptions over the past two, three decades. Nothing even comes close in terms of the revenue ramp up that we have seen from Anthropic. Speaker 6 (9:12) I mean, I think there are certain ways you can access some of these artificial intelligence sort of chatbots and systems at the moment. Obviously, Alphabet has its Gemini one. But do you think people are excited about Anthropic coming to the stock market? Because potentially it's the first sort of the new wave of them. I guess, you know, it's certainly beating OpenAI to the market. Speaker 1 (9:35) It is, but I would say SpaceX was one of the first ones. And SpaceX made a big pivot to AI right before the IPO, given how hot this whole AI infrastructure market has been this whole year. And so SpaceX, which also owns XAI, Speaker 1 (9:59) started to make that pivot towards renting their compute, their data center capacity to the likes of Anthropic and Alphabet. And they signed some large deals. And in the end, SpaceX IPO-ed at around a $2 trillion valuation. So by that yardstick, SpaceX... was an $18 billion company at the time of IPO, nothing even remotely close to what Anthropix growth rate is, which I said they grew their revenue sevenfold in the first nine months. And by that comparison, I mean, Anthropix IPO should definitely be north of $2 trillion. In fact, much higher if you apply the same multiple that SpaceX fetch in the market. Speaker 6 (10:46) Yeah, I mean, this is this, you know, everyone is talking about AI, you know, lots of excitement around it. But obviously, there's some quite big risks to consider. Certainly seen at the moment with people worrying about should there be a kill switch for AI systems or potentially just being far too smart? What are the key risks perhaps people should think about from an investment perspective? Speaker 1 (11:11) Yeah, I mean, look, the growth rates really matter when you are going public. And obviously with these companies, the cash burn is quite high. We have heard about OpenAI, in fact, kind of burning about 200 billion plus through 2030 in terms of their expense rate and all the funding they are using for deploying that data center capacity. Not a profitable business model if you aggregate everything from training to all the data center costs. Anthropic, on the other hand, has focused on enterprise from day one. They haven't really gone after that consumer use case, which OpenAI did, and now OpenAI is branching to enterprise. And so from that perspective, Anthropic should have a better profitability. kind of profile in terms of the gross margins, even though on an aggregate basis, if you sum up training and inferencing, these companies are losing money still. So it'll be interesting to see how that profitability kind of narrative evolves during the IPO and beyond. Speaker 6 (12:28) What about competition though? I mean, we're hearing lots about, you know, The fact that Anthropic seemed to be considered the best quality one, but now they're saying some of the other ones are catching up, particularly keeping an eye on what's happening in China. Do you think that people perhaps are underestimating competitive threats to Anthropic? Speaker 1 (12:50) Well, so open rate models have proliferated this year, and a lot of the open rate models are Chinese. I would argue, again, when you look at the numbers that Anthropic has talked about, you know,