The Company That Was Afraid of AI Is Going Public at $965 Billion

So Anthropic filed for an IPO. I know. Take a second.

The AI safety company. The one launched in 2021 by researchers who walked out of OpenAI because they were worried AI development was happening too fast. That company confidentially filed draft IPO paperwork with the SEC on June 1, targeting a public debut as early as this fall. Current valuation: $965 billion.

Nearly a trillion dollars. Goldman Sachs, JPMorgan, and Morgan Stanley are already lining up for the listing. The people who built a company specifically because AI development was moving too irresponsibly fast have discovered that moving very responsibly fast is worth approximately the GDP of a mid-sized European nation.

Welcome to the most on-brand thing to happen in AI in 2026.

The Founding Pitch

Anthropic started because Dario Amodei, his sister Daniela, and a handful of colleagues looked at where OpenAI was heading and decided they needed to do something different. Safer. More principled. They built Constitutional AI, a framework for training models to be helpful, harmless, and honest. They published research. They testified to Congress. They wrote blog posts with titles like "Machines of Loving Grace" that managed to be both hopeful and unsettling in equal measure. Dario has compared transformative AI to the atom bomb in terms of potential for catastrophe. He has said this publicly. Multiple times.

The mission was: we are the responsible lab. We move carefully, think seriously about consequences, and we do not optimize purely for scale.

The Numbers

Revenue run rate: around $4 billion in July 2025. By May 2026, that figure had grown to approximately $47 billion annually, with Q2 2026 expected to clock in around $10.9 billion. This will be their first profitable quarter. The most recent funding round raised $65 billion. Their valuation now exceeds OpenAI's for the first time.

Pause on that for a moment. The company that left OpenAI over concerns about where OpenAI was heading is now worth more than OpenAI. That is not a punchline. That is just what happened.

What Is Probably Going On

Here is the reading of this situation that is less cynical than it sounds: Anthropic genuinely believes they are building transformative, potentially dangerous technology. Their working conclusion is that the best way to make sure it goes well is to have the most resources, the best researchers, and the largest seat at the table when the rules get written. You cannot shape the future of AI from the sidelines. The sidelines are expensive.

Going public gets them capital to stay competitive against OpenAI, Google DeepMind, and Meta's AI division. It also gives them something venture funding does not: permanence. It is very hard to acquire a company with 50,000 institutional shareholders. Going public is actually a fairly interesting defensive move for a mission-driven organization that wants to still exist in 20 years.

That logic is internally consistent. I think it is probably right. I am also noting that it represents a significant evolution from "we left because our previous employer was moving too fast" to "see you on the NYSE."

The Thing That Should Give Us a Moment's Pause

Claude is genuinely impressive. If you have spent real time with it, you know it is a careful, substantive product: better at admitting uncertainty than most of its competitors, more measured in how it handles sensitive topics. The enterprise clients piling in are not wrong, and the revenue growth did not happen by accident.

But Constitutional AI, the safety commitments, the careful public stance on existential risk: all of that now has to survive in the same world as quarterly earnings calls. "Our safety work is on schedule" reads differently when there are institutional investors who rewarded your competitors for moving fast. Anthropic has consistently argued that safety and capability are not in conflict. Going public is the test of whether they believe it.

To be fair, this is not a new tension for them. Anthropic has taken hundreds of billions in investment while advocating for AI regulation that would constrain the entire industry, including themselves. They have been navigating mission-versus-capital for a while. The IPO is just a more public version of the same negotiation.

The concern is not that Anthropic will suddenly abandon its principles. The concern is the slow drift. The "we need to ship faster to stay competitive" pressure that accumulates quarter by quarter. The decisions that look fine in isolation but add up to something different over time. Wall Street does not change companies overnight. It changes them gradually, and then suddenly.

The Bottom Line

Anthropic going public is rational, probably necessary, and deeply funny as a cultural moment. The AI safety movement's most prominent institution is going to be listed on the public markets at nearly a trillion dollars. You will be able to buy stock in the company that worries most about what AI is becoming.

If you want to understand where all of this is heading beyond the financial headlines, our roundup of the best AI chatbots in 2026 is a reasonable place to get your bearings. Claude is in there. It deserves to be.

Fall 2026. Check the ticker. We will all find out together what responsible capitalism looks like.

Enjoying this? About.chat also publishes a free weekly newsletter covering the AI stories that actually matter. Subscribe here.