Anthropic plan would hand seven founders 50.1% of the vote on roughly 14% of the equity

Seven stakes of about 2% add up to roughly 14% of Anthropic ^[3][1]. Under the plan The Information described, that 14% would cast 50.1% of the shareholder vote on most matters ^[1][2], leaving the other 86% or so of the equity with 49.9% ^[2]. Per percentage point owned, a founder's shares would carry about six times the voting weight of anyone else's ^[6]. The founders would not pay for the difference: the new class raises their votes without enlarging their economic stake, according to the report ^[4]. At the $965 billion valuation of the May round, the stake they already hold was worth about $135 billion ^[8][3].

Board elections run on a separate track. Anthropic, a public benefit corporation, says its Long-Term Benefit Trust is an independent body whose members have no financial stake, and the trust holds its own class of stock with the power to elect and remove directors ^[6]. With the trust keeping a majority and the founders moving to three seats ^[5], seven seats less three less at least four leaves none for other shareholders unless the board grows ^[5]. One seat is vacant now ^[5].

Several terms are still open. Shareholders have to approve the class ^[10]. Founder control would hold only while at least three of the seven keep a minimum stake ^[2], a figure the report did not specify ^[10]. A separate employee class could break ties on certain matters ^[7]. The Information compared the arrangement to Palantir's founder-control structure ^[11].

At the $2 trillion valuation in preliminary Sep. 19 reports, a $100 billion raise would buy roughly 5% of the company ^[9][4]. If votes outside the founder class split in proportion to equity, that 5% would carry about 3% of the vote on most matters ^[7]. The SEC's IPO guidance says extra-vote shares can let founders control a company without owning most of its equity, leaving public shareholders with less influence ^[12].

I think the new class costs IPO buyers less than the 50.1% figure suggests. The trust already holds the power to elect directors and would keep a board majority ^[6][5], so public holders were not going to pick this board in any case. The counter-case is that founder votes would reach most other shareholder matters ^[2], leaving public holders outvoted on both tracks. The view fails if the prospectus sets the minimum founder stake low enough that three of the seven can keep control for as long as they choose.