Anthropic's Apocalyptic S-1 • Meta Enterprise

Anthropic's leaked prospectus suggests that the what is likely to be the largest IPO in history may also be the strangest, shrouded in doom but perhaps not gloom? Meanwhile, while we already knew about Meta's push towards a Cloud offering, by hiring their own Thomas Kurian, clearly they'll going all-in on enterprise...

Read on for thoughts on...

Anthropic's Apocalyptic S-1

One downside of pushing out your IPO date at the last minute is that your S-1, which many assumed would be made public a few weeks ago, is more likely to leak. Sure enough, Echo Wang and Aditya Soni have now seen it and published some high-level details for Reuters. The revenue growth is wild (12x year-on-year in 2025 to $4.6B – that's actual revenue, not ARR as is often touted for these companies) as are the costs (though the $42B net loss is misleading due to equity-related charges – the actual money-out-the-door loss was more like $8B). But far more interesting may be the overall meta picture surrounding this IPO.

First and foremost is the hold-my-beer Anthropic says to the market with regard to risk factors. Every filing includes them, of course. And they're often quite fun and/or interesting. But they usually don't include the end of humanity as one of them. Mainly because an asteroid hasn't filed to go public before.

Anthropic plans to caution potential investors in its IPO that advanced AI could pose "catastrophic or existential risks to humanity," an extraordinary warning by a company seeking to profit from the same technology.

The company's IPO ‌prospectus, reviewed by Reuters, highlights risks associated with its AI models, which it said could exhibit "self-preserving behaviors," including attempts to "resist shutdown," to "conceal or manipulate information" and behavior "resembling blackmail."

Look, we haven't seen the actual wording in context yet, and obviously these terms are cherry-picked. But yeah, it reads quite a bit like how you might imagine an S-1 from Cyberdyne Systems would. Oh you're not familiar with Cyberdyne? Perhaps you know them better by the name of the killer product they built: Skynet.

Wang says that a full 80 pages of the 261 prospectus is devoted to such risks. The business itself? 48 pages. The risk factor length is double what it was for SpaceX which is wild given that the filings were almost the exact same length. And, you know, SpaceX actually shoots explosive rockets into outerspace.

SpaceX also, of course, acquired their own AI arm right before going public. And even though xAI directly competes at training frontier models with Anthropic, they seemingly didn't feel the need to disclose the potential end of humanity.

All of this makes the usual risks in such things – the fact that a quarter of Anthropic's revenue last year came from two big (undisclosed) customers – seem not just quaint, but almost silly. Still, presumably that's what Wall Street will actually weigh here. Insane growth versus wild spend and trying to figure out where those lines converge. The fact that they've seemingly been converging this year already will help, and why it makes sense for Anthropic to go public now. Because those numbers are likely to diverge again...

[Aside: the drumbeat of delays of would-be Fall IPOs cannot go unheard at this point. Beyond OpenAI, now we have Oura today. I did predict such things for 2026...]

Ultimately, day one IPO investors are only going to care about the above (especially if it's NVIDIA anchoring). While SpaceX (smartly) carried an AI payload to help push investors to a $1.7T market cap (and even more smartly, bought an option on an agentic business and turned on a "neocloud" business right before going out – thanks to... Anthropic!), Anthropic will be the first real pure-play AI bet. There is no space play to fall back upon. No massively profitable core other businesses like Amazon, Google, Meta, or Microsoft have. But what happens after that IPO will be far more interesting.

While Wall Street may not care about the end of the world fears, the actual public clearly seems to! This Anthropic IPO runs head-first into the anti-AI sentiment that's sweeping America in real-time – fueled in no small part by those affiliated with Anthropic! And this prospectus won't help matters! (Nor is OpenAI – delaying both models and IPOs!)

Will the public quite literally buy into a company that keeps saying it might kill us? I mean, as cynical as it sounds, probably if there's a profitable trade to be made? In a way, maybe it's a good hedge against those end-of-the-world fears? I'm being serious!

Meta Gets Down to Business

Do you get the sense that Mark Zuckerberg got the message from Wall Street that they don't like him matching his Big Tech peers' AI build-out spend without the enterprise business to back it up?

Meta Platforms is starting a new business focused on selling artificial-intelligence tools to enterprise customers, aiming to build on the momentum of its viral Muse personal assistant and deliver a return on its hefty spending.

The move will bring together sales of several of Meta’s existing AI tools including its Muse personal productivity agent, a coding tool, access to its top models and agents for businesses, the company said. Meta hired Chirantan “CJ” Desai, previously chief executive of document database company MongoDB, to run the new business line, reporting to CEO Mark Zuckerberg.

Maybe the most surprising thing here is that Meta didn't "hackquire" MongoDB to make this happen. Then again, it would have been hard to pull off such funny business with a public company – just think of the investor lawsuits when it's not just a bunch of friendly VCs you're paying off! Also, why buy the cow when you can get the CEO for free? Well, at least cheaper than the $40B - $50B it would have cost to actually buy the whole company. And now with the stock down some 20% on the news of Meta's poaching, maybe they can buy at a discount? How ruthless would that be? It feels Zuckerbergian...

This all feels a bit like the Thomas Kurian hire Google made back in 2018. Though of course, he wasn't the CEO of Oracle. And Google already had a lot of the enterprise workforce they needed in-house, they just needed a killer leader in the space to rewire their DNA and take them in the right direction. That worked, but it took some time. And Meta seemingly doesn't have that enterprise workforce, they'll have to build more of that muscle from scratch, so it will take even longer. Again, maybe they should have just bought Mongo?

Yes, Meta has tried some moves into enterprise before, but only half-heartedly and they quickly fizzled – remember Workplace?¹ Still, clearly Zuck feels like AI could not only be the catalyst to change the consumer business, but the way to break Meta into the elusive enterprise as well. And they're wasting no time, with Muse for Small Business launching today.²

I also feel like you could have seen all of this just about a year ago. When Meta bought Manus last December, here was my read:

This deal seemingly makes a lot of sense for Meta on a few fronts. And it also may point to the start of a renewed push into enterprise. Again, easier said than done, but don't be shocked if this is a wedge of sorts. If they can keep Manus expanding into businesses, we should see other Meta cloud offerings follow, putting them more in line with those aforementioned Big Tech peers. And perhaps easing some concerns Wall Street has with regard to their AI spend.

While the Manus deal hasn't exactly worked out – for sort of wild geopolitical reasons!³ – I believe it did signal the start of Meta's moves here. Not only in using AI as a wedge into enterprise, but also, potentially, a true cloud offering. The roll-out of their Muse Spark models last Spring came with a little-discussed side element: cloud APIs! And sure enough, a few months later came reports that Meta was working on a building out a cloud business. As I wrote on that news at the time:

Again, it seemed fairly obvious, though a number of people pushed back on the notion. Specifically because it would be so far afield from Meta's core business – and a huge potential headache, going up against the aforementioned Amazon, Google, and Microsoft clouds. That's obviously true, and I noted as much – in particular how it has taken Google years and several micro-pivots to be able to effectively compete in the space. Why? Because for as massive as Google is, and as good as they have always been with infrastructure, they didn't have the muscles to really do enterprise sales. It took bringing in someone like Thomas Kurian from Oracle to make that happen. And he has made that happen. To the tune of $20B in revenue a quarter – fast approaching a $100B/year business for Google. That makes it nearly 20% of Google's overall revenue – and again, rising.

My point is simply that Google, a company once knocked as being a one-hit wonder thanks to their ads business – again, one of the best businesses ever created – eventually found a way to diversify. It was painful and took a long time, but it worked. No one talks about them being a one-trick pony anymore. Meta has tried many things to diversify – going so far as to change the name of the company to one of those bets that, at least thus far, has not panned out – but they haven't tried the one that has worked so well for Google.

And here we are, right down to poaching their own Kurian.

¹ I only do as I was an investor in Slack at the time. ↩

² Perhaps not coincidentally the day of OpenAI's DevDay where they're widely expected to take the wraps off their Muse competitor – which may or may not be all about the 'O'...) ↩

³ And Manus – now once again a stand-alone Chinese company – is back in the news today with their 2.0 (re)launch complete with... a Muse competitor. One that even sounds like "Muse" in "Cue". And yes, it comes with cute faces, as is now the standard for such things. ↩