Anthropic's Founders Grab Super-Voting Shares Ahead of a Potential $2 Trillion IPO

As Anthropic races toward what could be the largest IPO in history, the company is shoring up the one thing money can’t easily buy on the public markets: control. According to a report by The Information confirmed by Reuters on August 18, 2026, Anthropic has been preparing to grant CEO Dario Amodei and its other co-founders a class of stock with outsized voting power — a “supervoting” structure designed to insulate management from shareholder pressure long after the company lists.

The move is the latest piece in an elaborate governance architecture that Anthropic has been assembling since its founding, and it sets up a striking tension at the heart of the Claude maker’s public debut: the same company that has staked its brand on safety-first, mission-driven AI is preparing to sell shares to public investors who will have essentially no say in how it is run.

Founders with 2% of the equity, most of the power

The numbers explain the motivation. Dario Amodei reportedly holds only around 2% of Anthropic’s equity — a modest stake for a founder-CEO of a company investors now expect to float at $2 trillion or more. Under a conventional single-class structure, that 2% would translate into negligible voting influence the moment the company goes public and new shares flood the market.

A dual-class structure with supervoting founder shares solves that arithmetic problem. Much as Mark Zuckerberg used a similar arrangement at Meta to retain control with a minority economic stake, Amodei and co-founders — including president Daniela Amodei — would be able to direct the company’s strategy regardless of how many public investors buy in. Reports suggest the founders’ voting control would be paired with restrictions that limit public shareholders’ influence over board composition.

For a company at the frontier of AI development, the rationale is familiar: insulating leadership from quarterly-earnings pressure is exactly what you want if your stated mission is the safe development of transformative AI. Critics of short-termism in public markets have long argued that frontier labs should not be hostage to index funds demanding cost cuts on safety teams.

The Long-Term Benefit Trust isn’t going anywhere

The supervoting shares are only half the story. Anthropic also plans to retain the Long-Term Benefit Trust (LTBT), its most unusual governance feature, according to the reports. The LTBT is a body of independent trustees — who hold no equity in the company — that holds special shares allowing it to elect a growing share of Anthropic’s board as the company hits capability milestones.

That structure inverts typical founder entrenchment. In most dual-class companies, founder control is a fixed shield. At Anthropic, the trust’s board influence increases over time as AI systems become more capable, on the theory that as the stakes of the mission rise, so should independent oversight. The trust can currently elect a minority of directors, with that power expanding along a pre-agreed schedule tied to progress on advanced AI.

Stack the two mechanisms together and the picture for IPO buyers becomes unusual: founders vote like giants thanks to supervoting stock, and an independent trust controls board membership independent of any shareholder — public or private. As one analysis put it, public investors in the IPO would get economics, but no meaningful path to control.

A $2 trillion price tag on a governance experiment

The timing is not accidental. Investors told the Financial Times they expect Anthropic to float at a valuation of $2 trillion or more, potentially as soon as October 2026 — a figure that would eclipse every IPO in history, including SpaceX’s record-setting debut. The company reportedly targets $190–200 billion in revenue for 2028, up from an annualized run rate of around $47 billion disclosed in May, with Fable 5 and Mythos 5 driving enterprise adoption.

At $2 trillion, Anthropic would trade at roughly 42 times its current run rate — a multiple that only makes sense if you believe the revenue trajectory. And that is precisely where governance matters. The bull case for the stock is that Anthropic’s leadership can keep making aggressive, long-horizon bets on frontier models without being second-guessed. The bear case is that the same insulation means nobody — not founders’ fellow shareholders, not the market — can force a course correction if those bets fail.

Not everyone on Wall Street is buying the headline number. Reports indicate Goldman Sachs and JPMorgan, the banks reportedly leading the offering, are modeling valuations closer to $400–500 billion — a striking gap versus investor expectations, and a reminder that the $2T figure is aspirational pricing from existing backers rather than a settled number.

What it means

For the AI industry, Anthropic’s IPO structure is becoming a template test. If the market accepts a $2T valuation on a company where public shareholders have no board control, founders everywhere will take note: mission-driven governance and founder dominance can coexist at scale. If the banks’ more conservative models win out, it will signal that investors still price governance risk into even the hottest AI assets.

For Anthropic itself, the supervoting preparation is essentially the final lock on the door before going public. Between founder super-votes and an ever-strengthening trust, the company has ensured that whoever owns the shares, the people who run Anthropic — and the trustees who watch them — will remain the ones steering the most closely watched safety experiment in commercial AI.

The IPO window is reported for late 2026, with October the most frequently cited target. Whenever it lands, one thing is already clear: buying into Anthropic won’t buy you a say in Anthropic.