Business
Anthropic Just Filed for IPO at $965B. The Psychology of the AI Gold Rush.
Anthropic's S-1 filing, $47B revenue, and a $965B valuation. Three AI giants racing to IPO in the same quarter. This isn't just business news — it's herd behavior at civilisational scale.
Three companies. $3 trillion in combined valuation. One quarter.
On June 1, 2026, Anthropic — the company behind Claude — confidentially filed its S-1 with the SEC. Valuation: $965 billion. Revenue run-rate: $47 billion, up from $10 billion a year ago. That's 5x growth in twelve months.
Ten days earlier, SpaceX filed its public S-1 for a June 12 Nasdaq listing. OpenAI filed around May 22, targeting September at $1 trillion+.
Three of the most consequential companies on the planet, all racing to go public within weeks of each other.
This isn't a coincidence. It's a behavioral pattern. And it's worth decoding.
The Herd Effect in Capital Markets
Why would three companies file in the same quarter? Each could have waited. Each had enough private capital to stay private for years.
The answer is informational cascading — a behavioral economics concept where each actor's decision is influenced by observing the previous actor's decision, regardless of their own private information.
SpaceX filed first. That created a signal: the IPO window is open. Anthropic filed within days. Not because their business fundamentals changed overnight, but because the signal changed. If SpaceX thinks the market is ready, maybe it is. If Anthropic files too, OpenAI can't afford to wait — being last signals weakness.
Each filing makes the next one more likely. Each one validates the others. This is the herd effect operating at the highest level of corporate finance.
The irony? The companies building AI to reduce human irrationality are being driven by the most ancient form of it.
The $965 Billion Question
Let's sit with Anthropic's number for a moment. $965 billion.
That's more than the GDP of Saudi Arabia. For a company that didn't exist five years ago. That makes a chatbot. A very good chatbot — but a chatbot.
How do you justify that number? You don't. Not with current revenue. Not with current margins. You justify it with narrative.
Anthropic isn't being valued on what it earns today. It's being valued on the story investors tell themselves about what AI will be worth in 2030. And that story is compelling: AI will be everywhere, Claude will power enterprise workflows, agentic AI will replace entire departments.
Maybe that's true. Maybe it isn't. But the valuation doesn't reflect probability-weighted outcomes. It reflects anchoring — once someone says "$965 billion" with a straight face, that number becomes the reference point. Every subsequent analysis adjusts from that anchor, not from zero.
This is how bubbles form. Not through stupidity, but through anchoring.
The Dot-Com Mirror
Every generation thinks their bubble is different. It never is.
In 1999, the story was: the internet will be everywhere. And it was true — the internet did become everywhere. But that didn't stop 78% of dot-com companies from going to zero. The thesis was right. The valuations were wrong.
In 2026, the story is: AI will be everywhere. And it probably will be. But $47 billion in revenue supporting a $965 billion valuation means the market is pricing in 20x revenue for a company in one of the most competitive markets in history.
Google has Gemini. Meta has Llama. Apple has its own models. Microsoft has its OpenAI partnership. Amazon has Bedrock. Every big tech company is building AI. And open-source models are getting dangerously close to frontier performance at a fraction of the cost.
The competitive moat for any AI company is thin and getting thinner. But the valuations keep climbing.
This isn't analysis. It's availability bias — investors overweight the vivid, recent success stories (ChatGPT going viral, Claude becoming essential for developers) and underweight the base rate of how competitive markets actually work.
FOMO as a Business Strategy
Here's the part nobody says out loud: these IPOs aren't really about raising capital. Anthropic just raised $65 billion in Series H. They don't need public market money.
The IPO is about liquidity for early investors and employees, credibility for enterprise sales ("we're a public company, we're not going anywhere"), and most importantly, timing.
The fear isn't that the market won't value them highly. The fear is that the window closes. That regulation hits. That a price war with Google crashes margins. That the next model generation disappoints.
Filing now isn't confidence. It's FOMO. The same FOMO they'd diagnose in their users.
What This Means for You
If you're watching this as an investor, an employee considering AI startups, or someone building a career around AI — here's the decode:
The technology is real. The applications are real. The revenue is real. But the valuations are a story. Stories can change overnight.
The people who made money in the dot-com era weren't the ones who bought at peak excitement. They were the ones who understood which companies had actual moats (Google, Amazon) and which were narrative plays (Pets.com, Webvan).
In the AI era, the same filter applies. Which companies have something that can't be replicated? Which ones are being valued on what they are versus what we hope they'll become?
The Decode
Anthropic's S-1 isn't just a financial filing. It's a behavioral event.
It triggered informational cascading across capital markets. It anchored a $965 billion number in everyone's mind. It activated availability bias in investors who watched ChatGPT transform their workflows. And it created a FOMO-driven IPO race that may define — or destroy — the next decade of AI.
The smartest people in the world are building AI to make humans more rational. And they're going public using every irrational impulse in the book.
If that doesn't make you think, nothing will.