What We Know About Anthropic's $250M Founder Payouts in IPO

Anthropic's S-1 registration statement discloses that each of the company's seven co-founders will sell exactly 3,378,378 shares in the initial public offering, a figure that produces roughly the same cash payout for every founder regardless of title or role, according to TechCrunch.

How much will each founder sell and receive?

Anthropic has set an initial price range of $70 to $78 a share, TechCrunch reports. At the midpoint of $74, each founder's 3,378,378 shares would convert to just under $250 million in cash. At the low end of the range, $70 a share, each founder would collect just over $236 million; at the high end, $78 a share, the payout would run just under $264 million. Companies frequently price above their stated range when demand is strong, and the S-1 is already showing signs of that: board member Reed Hastings has indicated he wants to buy $300 million worth of shares through his donor-advised fund, and board member Christopher Liddell has signaled interest in buying $3 million worth, per TechCrunch's reporting on the filing.

Who are the seven founders selling shares?

The sellers named in the filing are Dario Amodei, chief executive; Daniela Amodei, president and board chair; Tom Brown, chief compute officer; Jack Clark, head of public benefit; Jared Kaplan, chief science officer; Sam McCandlish, chief architect; and Christopher Olah, interpretability research lead. All seven left OpenAI in 2021 to start Anthropic after disagreements with OpenAI's leadership over how artificial intelligence should be built, TechCrunch notes, with siblings Dario and Daniela Amodei — who had held senior research and policy roles at OpenAI — joined by five former colleagues. Anthropic has said since its founding that it is more focused on AI safety than its rivals, including the company the founders left.

What happens to each founder's stake after the sale?

Per the S-1, each founder will own about 3.2% of total outstanding Class A shares after selling, or roughly 22% collectively. That figure describes Class A ownership only, not the founders' share of the company overall. Anthropic's capital structure includes Class N shares, which are nonvoting and held by investors including Amazon and Google. Because those Class N shares exist alongside Class A, the founders' actual stake in the company's total equity is considerably smaller than the 22% Class A figure suggests, TechCrunch reports.

How do the founders retain control if they're selling shares?

Each founder will also receive one share of Class F stock, which carries supervoting rights, according to the filing. Collectively, the seven will control 51.1% of total voting power after the IPO — enough to keep the group in charge of the company even as they sell down a meaningful slice of their economic stake. The filing specifies that no single founder will hold significantly more shares or votes than the others, preserving an even split of power among the seven rather than concentrating control in any one executive.

IPO Numbers at a Glance

What to watch as the IPO proceeds

• Final pricing — Anthropic's $70–$78 range is wide for a company reportedly targeting a $1.4 trillion valuation; watch whether strong demand from investors like Reed Hastings pushes pricing above the stated range.
• Hastings' donor-advised fund purchase — the $300 million order flagged in the S-1 would be among the largest individual commitments disclosed so far.
• Class N shareholder behavior — Amazon's and Google's nonvoting stakes mean their economic interest in Anthropic is separate from any governance influence, a distinction likely to draw scrutiny as the offering closes.
• Equal-split structure — because all seven founders are selling identical share counts and retain equal Class F supervoting shares, any future change to that balance would mark a departure from the structure set out in this filing.

Why did Anthropic's founders leave OpenAI?

TechCrunch's reporting frames the departure as a matter of public record: the seven left OpenAI in 2021 over disagreements about how AI should be developed, then built Anthropic around a publicly stated emphasis on AI safety relative to competitors, including OpenAI itself. The S-1 does not elaborate on those disagreements, and no further detail on the split appears in the filing material reviewed.