Anthropic, rejected by 21 top-tier VCs: The most expensive "miss" in AI history

Revisiting this fiasco, we only saw one word: Serves you right!

By: SynTelligence

Editor’s note: Who would have thought that when OpenAI’s core team set out to start a company, they were turned away by 21 top-tier VCs? Five years later, these same people, just to get a seat at the table, were willing to pay a 300x premium.

In 2021, Anjney Midha took Anthropic’s BP to 22 meetings with top-tier VCs—only to be kicked out 21 times.

Fast-forward to January 2026: Anthropic’s latest round of financing brings in $25 billion, and the valuation immediately jumps to $350 billion.

What does that even mean? It’s like 10 instances of OpenAI in 2023.

Back then, those investment heavyweights who welded the door shut under the pretext of “risk control” are probably standing in line in the restroom, crying and passing out.

This isn’t “getting slapped.” It’s the most expensive collective “IQ tax” incident of this century.

21 rejection letters: the “blind-eye” moment for top-tier VCs

All those who rejected Anthropic were “heroes” in Midha’s eyes back then.

Look at Anthropic’s lineup at the time: OpenAI’s core team’s breakaway executives—plus the biological parents of GPT-3.

With this setup today, it’s the level of “the money arrives before the PPT is even finished.”

Midha thought this time was a lock—then reality slapped him in the face.

Back in 2021, in the eyes of VCs, large models were just a bottomless pit that ate money.

On top of that, that group at Anthropic was almost obsessively fixated on “AI safety,” even setting up a nonprofit background. Mainstream VCs back then simply didn’t get it, and traditional capital directly labeled them as a “high-risk oddball.”

It wasn’t until Spark Capital led the Series C that this crew finally woke up. Later, Jason Shuman had to admit:

Turns out, projects that everyone could understand early on usually don’t have much in them.

How big was the cost of this bout of “cognitive sluggishness”?

In May 2021, Anthropic accepted a $124 million Series A led by Jaan Tallinn.

Compared with today’s $350 billion valuation, those 21 rejecting institutions back then missed out on a return of nearly 3000x.

Risk control—this is the biggest risk

In this drama, Sequoia Capital perfectly embodies what it looks like to “lose its bearings.”

According to reports, back then, Sequoia’s global decision-maker, Roelof Botha, repeatedly declined to lead early rounds.

The reason sounded especially lofty: “Concentration risk.” In other words, they were afraid of getting hung up on a single AI tree, and that it would disrupt the balance of asset allocation.

These are the “correct” buzzwords of traditional finance—yet in the face of AI’s exponential growth, they’re basically a disaster.

Sequoia only turned around once its face had been beaten up. By early 2026, the actual contribution rate of AI investment to the U.S. GDP had surged to 40%.

At that point, who still talks about allocation? This is life-or-death assets! So Sequoia’s management team got a major reshuffle. After Alfred Lin and Pat Grady took the stage, they quickly overturned Botha’s conservative dogma.

Roelof Botha publicly responded to the leadership changes at Disrupt in 2025, and defended Sequoia’s “free speech” culture

In January 2026, Sequoia finally squeezed its way, biting the bullet, into Anthropic’s latest round.

The awkward part is that by then, the valuation had already skyrocketed from $1 billion in the Series A to $350 billion.

For the so-called “risk avoidance,” Sequoia sat outside doing nothing for 5 years—and in the end, with tears, paid a “cognitive premium” over 300 times.

This isn’t just Sequoia’s problem. The data back then was brutal:

Before Spark Capital entered, most VCs would rather invest in those harmless SaaS software products and wouldn’t dare touch Anthropic, which burns tens of billions of dollars in compute every year.

Compared with “investing the wrong way,” what these people feared even more was being the “early bird,” only to end up as a punchline “bare swimming” in the tide of the era.

The dimensionality-reduction strike from “non-mainstream” capital

While mainstream VCs were still calculating ROI, who saved Anthropic?

A group of “madmen.”

Jaan Tallinn, who led the Series A in May 2021, is the co-founder of Skype and also an ardent believer in AI safety. He completely upended Wall Street’s money-throwing logic:

“I’m investing money not to profit from large model gains. I’m afraid AI will run out of control and get humans killed.”

His logic was “capital substitution.” Using money that cares about whether humans live or die to squeeze out all that copper-smelling