For months, the financial world has been watching and waiting for Anthropic to make its move, and Monday morning it finally arrived: the artificial intelligence juggernaut formally filed its prospectus for an eagerly anticipated initial public offering. The document was more than just a bureaucratic step toward the public markets; it was the company’s first real, hard look under the hood, revealing details that had never before been made public. Among the most striking disclosures were the ownership stakes of Anthropic’s seven cofounders, a group of former OpenAI researchers who left in 2021 to build what they hoped would be a safer, more ethical AI company. According to Forbes’ calculations, each cofounder holds roughly a 1.4% stake in the company, with individual estimated values ranging from $13.1 billion for chief architect Sam McCandlish, 36, to $13.8 billion for chief compute officer Tom Brown, 39, and interpretability research lead Chris Olah, 34. The numbers are breathtaking, especially when you remember that these were relative unknowns in the tech world just a few years ago, researchers toiling in a field that many outside Silicon Valley barely understood. Now, they are each sitting on paper fortunes that rival the GDPs of small nations, all based on Anthropic’s $965 billion valuation from its last private funding round in May. It is a dizzying leap, and yet it may only be the beginning.
The naming of those fortunes comes with a few psychological wrinkles, because not all cofounders are equal in the public imagination, even if their net worth numbers are remarkably close. Anthropic’s two most high-profile cofounders, CEO Dario Amodei, 43, and his sister, president and chairperson Daniela Amodei, 39, are estimated by Forbes to hold stakes worth $13.6 billion and $13.7 billion, respectively. Their other cofounders—head of public benefit Jack Clark, 38, chief science officer Jared Kaplan, 42, plus McCandlish, Brown, and Olah—all fall within the same narrow band of roughly $13.1 billion to $13.8 billion. That kind of uniformity is striking and speaks to a founding group that deliberately kept things equal, at least at the beginning. But the prospectus also revealed something more personal: the Amodeis were among the only cofounders who participated in share sales during recent tender offers. In April 2026, executives, employees, and investors sold a combined $4.7 billion in shares, and in June 2025, another $157 million changed hands. Alongside the Amodeis, notable sellers included CFO Krishna Rao, who didn’t even join the company until 2024, and Holden Karnofsky, who happens to be Daniela Amodei’s husband and a prominent AI philanthropy figure in his own right. The exact amounts each individual sold were not disclosed, leaving some tantalizing gaps in the story. Still, the fact that the Amodeis took some money off the table while their peers did not might say something about their personal financial planning, their confidence in the company’s future, or simply their different life circumstances. What is clear is that the IPO itself is likely to dwarf all previous private market milestones. Anthropic has not yet set a price range, but reports suggest it is seeking a valuation between $1.5 trillion and $2 trillion, which would push each cofounder’s net worth well past the $20 billion mark, an unfathomable sum for people who were once just working on research papers and training runs.
For all the eye-popping wealth on display, the prospectus also highlighted a deeply unusual and almost utopian aspect of Anthropic’s founding ethos: every single cofounder has pledged to dedicate 80% of their personal equity in the company to philanthropic causes. That is not a small gesture; it is a declaration that the people who built this company do not intend to keep most of what they create, at least in a material sense. But the fine print matters. The prospectus is careful to note that these commitments are revocable and not legally binding, which raises a fascinating question about the gap between intention and obligation. A pledge made in the quiet early days of a startup, when the outcome was deeply uncertain and high-minded ideals felt noble, may look very different when real billions are on the line. There is no legal mechanism forcing the cofounders to follow through, no contract that will compel them to hand over the bulk of their wealth if their priorities change. At the same time, some concrete steps have already been taken. The prospectus reveals that two donor-advised funds, which are commonly used vehicles for charitable giving, currently hold Anthropic shares worth around $260 million based on the May funding round valuation. It is not entirely clear how much of that stock came from each cofounder, but Dario Amodei has sole voting power over those particular shares, meaning he has the final say in how they are directed. That gives at least a small tangible shape to the philanthropic promise, even as the larger, loftier pledge remains a matter of trust rather than legal duty. It is a strange and thoroughly modern tension: a company built around the idea of protecting humanity from AI harm, whose own leaders are sitting on personal fortunes that could fund countless charitable projects, yet whose grandest commitments are ultimately as fragile as a change of heart.
The governance structure disclosed in the filing is perhaps even more unusual than the philanthropy, and it goes to the very heart of what Anthropic claims to be. Despite owning only about 10% of the company’s stock as a group, the seven cofounders jointly control a single non-economic share that wields 50.1% of the shareholder votes. In practical terms, that means the founders can remain firmly in charge even after the company goes public, regardless of what outside investors, newcomers, or activist shareholders might prefer. It is a powerful mechanism for retaining control, and it speaks to a deep desire to protect the company’s original mission from the pressures of the quarterly earnings cycle. But even that is not the final layer of the architecture. There is also an entity called the Long-Term Benefit Trust, which Anthropic describes as an independent body of financially disinterested trustees with expertise in AI safety, national security, public policy, and social enterprise. This trust has the power to elect and remove a majority of Anthropic’s directors, meaning that it, not the founders themselves, holds the ultimate reins when it comes to the boardroom. In many ways, this is an attempt to institutionalize a certain philosophy: the company should not simply maximize shareholder value; it should serve a broader human purpose. Whether it actually works in practice is another matter entirely. Plenty of companies have created “benefit corporation” structures or independent boards with noble intentions, only to see those mechanisms become symbolic over time. Anthropic’s structure, however, is more intricate than most, and it reflects a genuine intellectual conviction among the founders that AI is too dangerous to be governed solely by profit-seeking capitalists. The real question is whether this elaborate apparatus will survive the transition to major public ownership, when the pressure to deliver returns to a vast new class of shareholders becomes immense.
The story of how these seven people ended up in this position is now well known in tech circles, but the prospectus brings it into sharper focus. Anthropic was founded in 2021 by seven former employees of OpenAI, the massive artificial intelligence company cofounded by billionaires Sam Altman and Greg Brockman. The split was not just a matter of career ambition; it was rooted in a genuine philosophical disagreement about the direction of AI development. The founders left in part because they believed OpenAI was becoming too commercial, too focused on speed and market dominance, and not enough on safety and ethical considerations. They wanted to build a rival lab that would take AI risk seriously, and they gave it a legal structure designed to reflect that commitment. Since then, both companies have grown into colossi, and their rivalry has become one of the defining dramas of the tech industry. OpenAI, for its part, was valued at $852 billion in its most recent private funding round in March, and is reportedly raising even more money at a staggering $1.25 trillion valuation. It is also widely expected to file for its own IPO in the near future, setting up a head-to-head contest between the two former colleagues turned bitter rivals. The irony is almost too perfect: a group of people who left OpenAI to build a safer, more thoughtful company are now on the verge of an even more dramatic public spectacle, competing directly with the very institution they walked away from. Their success validates the original bet to some degree, even if the path has been littered with compromises, scaling challenges, and the sheer difficulty of building a business while remaining true to a mission that was always somewhat in tension with rapid growth.
In the end, the IPO filing is just the beginning of a long and uncertain journey. Anthropic declined to comment when reached, and the story is still developing, with many crucial details yet to come, including the actual IPO price, the final valuation, and the finer points of the share structure. But even at this early stage, the filing tells a remarkable story about the intersection of technology, wealth, and idealism. It is a story of people who wanted to change the world and ended up becoming extraordinarily rich in the process, who tried to build safeguards into their own company only to find that no amount of legal engineering can fully prepare you for the temptations of trillion-dollar-scale riches. It is also a story about the broader AI moment we are all living through: entire industries are being reshaped in a matter of years, fortunes are being created at a pace that would have seemed impossible in any previous era, and the consequences of these tools are so profound that even the people building them are not entirely sure what will happen next. Anthropic’s cofounders have pledged to give away 80% of their equity, they have structured their company so that independent trustees can override even the founders themselves, and they have placed their trust in a non-economic share that enshrines their original vision. Whether those safeguards hold up under the glare of the public market is impossible to know. What is certain is that the world will be watching closely as Anthropic makes its debut, not just because of the money, but because the outcome will say a great deal about the future of artificial intelligence, the responsibilities of the people creating it, and whether a company built on such high-minded principles can survive contact with the hard realities of Wall Street. For seven cofounders who once walked out of OpenAI to pursue a more careful path, it is finally time to face the world and prove whether their extraordinary experiment truly works.
