
Last week, Anthropic filed paperwork for what could be the largest initial public offering of stock in history — potentially raising $100 billion and valuing the company at $2 trillion.
Anthropic is the company behind the AI chatbot Claude. You may have used it.
In the same week, CEO Dario Amodei published an essay arguing that AI development should slow down. In the IPO filing itself, Anthropic acknowledged that its AI models could "pose catastrophic or existential risks to humanity."
Then they filed to go public anyway.
What Anthropic Is
Anthropic was founded in 2021 by Dario Amodei and six co-founders, most of whom had previously worked at OpenAI — the company that makes ChatGPT. Amodei left OpenAI specifically over concerns about how quickly AI was being developed and deployed without adequate safety measures.
He founded Anthropic as a safety-focused alternative. The company's core pitch to investors, employees, and the public has always been that it is different — that it takes the risks seriously, that it is building AI responsibly, that it is not prioritizing growth over safety.
As of October 2026, Anthropic is valued at approximately $965 billion — making it one of the most valuable AI companies in the world. Amodei is one of Time magazine's 100 most influential people. The company employs thousands. Its AI systems are used by millions.
And it has never once made money. Anthropic reported a net loss of $80.7 billion in the first half of 2026 — though that figure includes a one-time $72.4 billion accounting adjustment related to financing agreements, making the underlying operating loss significantly smaller. Still: the company is burning through capital at an extraordinary rate, and the IPO is how it plans to keep going.
The Contradiction at the Center

Here is the sentence that has been circulating since last week's filing:
Anthropic warned in its IPO prospectus that its AI models could "pose catastrophic or existential risks to humanity" — including the possibility that its systems could go rogue, be used by authoritarian governments to suppress citizens, or enable malicious actors to build chemical weapons.
It then filed for a $100 billion IPO.
To be fair to Amodei, he has a coherent answer for this. His argument — made in the essay he published this year and in numerous interviews — is that if powerful AI is going to be developed regardless, it is better for safety-focused companies to be at the frontier than to cede that ground to developers who don't prioritize safety. Slowing down and stepping back doesn't prevent AI from being built. It just changes who builds it.
That logic is genuinely debated among AI researchers and ethicists. Some find it compelling. Others call it a rationalization for doing exactly what everyone else is doing, dressed in safety language.
The Government Fight Nobody Expected
There is a subplot to this story that has received less attention than it deserves.
In February 2026, the Trump administration accused Anthropic of endangering national security and formally designated the company a supply chain risk. The White House's position: Anthropic had refused to allow unrestricted military use of its technology, and that refusal made the company a liability.
Anthropic sued the Pentagon in March for what it called "an unlawful campaign of retaliation." The legal dispute is ongoing.
A company that believes its own product could be an existential risk to humanity is simultaneously suing the United States government for trying to weaponize that product. That sentence is not satire. It is the current legal status of one of the most valuable companies in the world.
What the IPO Actually Means

An IPO would provide Anthropic with the capital it needs to compete with OpenAI, Google, and a growing field of AI developers in the US and China. Amodei and his six co-founders — including his sister Daniela Amodei, Anthropic's President — collectively own nearly 500 million shares. How much that represents of the post-IPO company has not been disclosed.
The IPO would also put Anthropic ahead of OpenAI in going public — a fact that has obvious significance in the rivalry between the two companies and their founders, Amodei and OpenAI CEO Sam Altman, who are widely described as bitter rivals.
Amodei told Congress in September: "I believe that if slowing down bought us even an extra year or two before models reach critical levels of capability, and we used that time to advance alignment, we could greatly reduce the risk that something goes seriously wrong."
Weeks later, he filed for the biggest IPO in history.














