Anthropic’s $900B valuation talks: the AI arms race is just getting started

If you thought the AI funding frenzy was cooling off, think again. Sources familiar with the matter tell me Anthropic is sitting on multiple pre-emptive offers at valuations between $850 billion and $900 billion, with the goal of raising a fresh $50 billion round.

Let that sink in for a second. $50 billion. That’s more than the GDP of half the countries on the planet. And the valuation? Nearly a trillion dollars for a company that didn’t exist a few years ago.

I’ve been covering this space long enough to remember when Anthropic was the scrappy safety-first alternative to OpenAI. Now they’re getting calls from investors who are basically begging to hand over cash before the round even formally opens. Pre-emptive offers at this scale tell me two things: first, the supply of high-quality AI startups is still way behind demand. Second, the big institutional players are terrified of missing out on the next OpenAI.

Anthropic’s last known valuation was around $60 billion after their Series E in early 2025. Jumping to $900B in roughly a year is a 15x multiple. That’s nuts by any traditional metric. But traditional metrics don’t apply here. These are bets on a future where Claude becomes the default reasoning engine for enterprise workflows, replacing everything from search to coding assistants to legal research.

What’s interesting is that Anthropic hasn’t even confirmed the round yet. The fact that they’re getting unsolicited offers at this level suggests the bankers are working overtime. I’d bet the final valuation lands closer to $1 trillion just for the headline. It’s marketing as much as fundraising at this point.

Of course, there’s the obvious question: can Anthropic justify this? Claude 4 is good, but it’s not 15x better than Claude 3. The real value is in the infrastructure and the moat they’re building around safety research. Investors are betting that their approach to alignment will become a regulatory requirement, giving them a defensible position that competitors can’t copy overnight.

But I’m skeptical. $50 billion is a lot of capital to deploy productively. Even if you burn $10 billion a year on compute and talent, that’s five years of runway. What happens if the market corrects or a competitor leapfrogs them? These valuations assume a winner-take-most outcome, and that’s a risky bet.

Still, I can’t blame the investors. If you’re managing a $100 billion fund and you think AI is the biggest platform shift since the internet, you write the check. You don’t haggle.

For now, Anthropic holds all the cards. They can pick their investors, set the terms, and keep building. Whether they can deliver on the promise is another story. I’ll be watching closely.