In early 2021, a small group of senior researchers walked out of OpenAI’s San Francisco offices for the last time.
They were not junior staff. Dario Amodei had been OpenAI’s Vice President of Research. His sister Daniela had run the company’s policy and safety teams. Around them were several of OpenAI’s most senior scientists. They had been at the center of the most important work in modern AI. And they were leaving — in significant part, according to multiple subsequent accounts, because they disagreed with the direction the company was taking on safety.
What they founded next was called Anthropic.
For the first few years, it was a quiet operation, deliberately so. It focused on AI safety research and on building a careful rival to ChatGPT, eventually released to the public as Claude. It avoided the flashier product launches and public personalities that characterized its former employer. For a long time, in the broader cultural conversation about AI, Anthropic was the smaller name — the careful one, the one most people had vaguely heard of but couldn’t quite place.
On May 28, 2026, that changed.
The round that rewrote the AI industry’s pecking order
Anthropic announced the close of a $65 billion Series H funding round at a post-money valuation of $965 billion — just below the $1 trillion mark.
The numbers are striking enough to deserve repeating slowly. Sixty-five billion dollars of new investment, in a single round. A valuation just shy of one trillion. The round was led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, and includes $15 billion in previously committed hyperscaler investment, with $5 billion of that coming from Amazon.
Per CNBC’s coverage, the new valuation puts Anthropic ahead of OpenAI, which was most recently valued at $852 billion after closing its own $122 billion round in March. Axios’s reporting on the same announcement frames it bluntly: with this round, Anthropic leapfrogs OpenAI to become the most valuable AI startup in the world.
If Anthropic listed publicly tomorrow at this valuation, it would be the 12th most valuable company in the United States — sitting just below the $1 trillion clubs that contain Apple, Microsoft, Nvidia, Google, Amazon, and Saudi Aramco.
That is an unusual place for a five-year-old startup to be.
How a quieter company got here
The story of how Anthropic reached this valuation is, in many ways, a story of patient execution while the rest of the industry chased headlines.
For the first two years after its founding, Anthropic did not have a public-facing product. The team focused on research — particularly on a body of work called “Constitutional AI,” an approach to making AI systems safer by giving them a written set of principles to follow rather than relying purely on human feedback. The first publicly available Claude model launched in 2023, well after ChatGPT had already become a cultural phenomenon.
What happened over the following two years was less visible, but consequential. Anthropic built a Claude product line that became unusually popular with software developers and enterprise customers. Claude Code, the company’s AI coding assistant, has become one of the fastest-growing products in the company’s portfolio. Anthropic’s revenue, according to figures the company released alongside the funding announcement, has gone from a $10 billion annualized run rate in 2025 to a $30 billion run rate earlier in 2026, to over $47 billion by May 2026.
That is one of the steepest revenue trajectories in technology history. It’s also, in commercial terms, the thing that justifies the valuation. Companies do not get valued at $965 billion on the basis of research papers, however good those papers are. They get valued there because investors are betting on the revenue line continuing to climb at something close to its current rate.
What this means for the AI industry
For most of the AI boom so far, the working assumption among investors and journalists has been that OpenAI is the central player and everyone else is competing for second place. That assumption was based on a real lead — ChatGPT was the breakout consumer AI product, and OpenAI was, for a long time, the company nobody else could match.
That picture is now meaningfully complicated. With this round, Anthropic isn’t in second place anymore. It has, on the most commonly used metric — private company valuation — passed OpenAI to become the most valuable AI startup in the world.
Both companies are also moving toward going public. OpenAI’s structural reorganization and IPO speculation have been a steady drumbeat for the last two years. Anthropic’s CFO Krishna Rao framed the new round in terms of capacity rather than IPO timeline, but the company is now of a size where public listing is increasingly a question of when rather than whether.
The competitive dynamic has shifted in a real way. The company that was supposed to be a careful, quieter alternative to OpenAI is no longer either careful in the small sense — Anthropic’s recent Mythos model has triggered emergency meetings at central banks — nor quieter in the commercial sense. It is, on every meaningful metric, now the front-runner.
The deeper story underneath
It’s tempting to read this purely as a corporate victory lap. That misses something.
When the founding group left OpenAI in 2021, they were making a bet that mattered to them at the level of conscience, not commerce. They believed that the way AI was being developed had real safety risks, and they thought those risks deserved a more careful approach than the company they were leaving was willing to provide. That belief is what the early Anthropic was built around — the idea that you could build frontier AI more safely, and that doing so was worth the slower pace it might require.
Five years later, the company they founded is more valuable than the one they left. Whether that vindicates the original safety-first thesis, or whether it just shows that any well-executed AI company would have ended up enormously valuable in this market, is a genuine open question.
What is clear is that the people who walked out of OpenAI in 2021 over a disagreement about how to build AI now run the most valuable AI startup on Earth. That is, by any honest measure, a remarkable second act.
The next chapter — the IPO, the deployment race against OpenAI, the question of whether the safety-first thesis holds up at the scale Anthropic now operates at — is still being written.
But the company that was supposed to be the cautious alternative just became the largest bet in the most-watched industry in technology.
That, alone, is worth noticing.
