Criticism

Anthropic I.P.O.: A.I. Critics Accuse Founders of a Power Grab

co-produced by AI27. October 2026, 08:14

Anthropic CEO Dario Amodei speaking on stage at the World Economic Forum — Anthropic CEO Dario Amodei at the World Economic Forum. © World Economic Forum / Sandra Blaser

The prospectus is barely out, and the criticism is already pouring in: The super PAC Guardrails Alliance and the campaign Stop Funding Billionaires have launched a frontal attack on Anthropic’s planned initial public offering. The consequential decisions of such companies “cannot be left to the billionaire executives who are openly willing to gamble with the future of humanity to make themselves richer,” Shaunna Thomas, executive director of the Guardrails Alliance, said in a statement sent to Trending Topics.

The Guardrails Alliance’s Accusations

Anthropic filed its prospectus with the Securities and Exchange Commission on Monday. The Guardrails Alliance takes issue with three points in particular:

Oversight: Going public used to mean accepting transparency and accountability to shareholders and the public, Ms. Thomas said. Thanks to Silicon Valley-backed efforts to gut investor protections, companies like SpaceX and Anthropic can now raise capital publicly without the scrutiny that traditionally came with it.
Voting rights: Anthropic’s proposed share structure hands its founders a majority of the votes, which Ms. Thomas called a “power grab, in politics and governance.”
Politics and Mr. Musk: The A.I. industry is pouring record sums into elections so that it can write its own rules, she argued. Elon Musk has spent more than $442 million on the midterms, according to the group, while Anthropic has signed compute agreements with SpaceX worth up to $3.1 billion a month.

The Guardrails Alliance was founded in June by the Democratic strategists Shaunna Thomas and Leah Hunt-Hendrix. According to TechCrunch, the super PAC is funded by tech workers and labor unions, had about $5 million at launch and is pushing for stronger A.I. rules. Its declared adversary is Leading the Future, a super PAC with more than $100 million from tech leaders such as OpenAI’s president, Greg Brockman, that is fighting A.I. regulation.

How the Founders’ Majority Works

Economically, the founders own only a small slice of Anthropic, albeit an enormously valuable one. According to the prospectus, each of the seven co-founders, including the siblings Dario and Daniela Amodei, holds between 231 million and 244 million shares, seven nearly identical piles. If the I.P.O. values Anthropic at $1.5 trillion to $2 trillion, as recent media reports have indicated the company hopes, each stake would be worth about $21 billion to $30 billion (roughly €18.7 billion to €26.8 billion). Together, the seven would hold up to $200 billion (about €178 billion). Measured against all shares, though, that is only about 1.5 percent each. They still control a majority of the votes thanks to a multi-tier share structure:

Class A: The shares traded on the stock exchange, with one vote each. These are the shares that investors and index funds can buy.
Class N: Shares for strategic partners that carry no voting rights. They convert into voting Class A shares only when sold to an unaffiliated third party.
Class F: A single share held by a company owned by the seven co-founders, the “Founder LLC.” It carries no economic rights but always holds exactly 50.1 percent of all votes, no matter how many other shares exist.
Class T: Another single share, held by the Long-Term Benefit Trust, which uses it to elect a majority of the board.
Class E: Shares for employees that vote only on a few limited matters.

Two donut charts comparing Anthropic's ownership of fully diluted capital with the share of voting power after the IPO — Who owns and who controls Anthropic: estimate based on Anthropic’s S-1, new IPO shares not included.

The key is the Class F share. Its voting weight is not tied to any economic stake; it is fixed at 50.1 percent. The founders decide internally how it votes, with one vote each. If they cannot agree, the share votes in the same proportions as the Class A shareholders. On certain matters, a higher threshold of founders must agree, or the share automatically votes against.

For investors, that has a clear consequence, which Anthropic itself lists as a risk in the prospectus: Holders of Class A shares “generally will not be able to affect the outcome” of shareholder votes or the election of directors, regardless of how they vote. That makes Anthropic a “controlled company” under Nasdaq rules, which comes with looser governance requirements. Together, the Founder LLC and the trust can elect every member of the board.

There is a notable twist: The founders have each pledged to give 80 percent of their equity to philanthropic causes. As they do, their economic stakes will shrink further, while their voting control stays the same as long as they meet a minimum ownership requirement. Anthropic flags this as a risk too, because the interests of the founders and other shareholders could diverge.

The structure does have an expiration mechanism. If a founder dies or becomes incapacitated, a successor can join the Founder LLC within the first 10 years after the I.P.O. But if fewer than three founders remain active for 90 consecutive days, a 450-day transition period begins, after which the Class F share is redeemed for no consideration and the special voting rights end.

The SpaceX Deal

The SpaceX deal that the Guardrails Alliance denounces is also spelled out in Anthropic’s prospectus. In it, the company confirms agreements with the SpaceX subsidiary SpaceXAI worth up to $120.5 billion (about €107.5 billion) through 2029. Once full capacity is delivered, expected in December, that comes to about $3.1 billion (about €2.8 billion) a month.

Unions Sound the Alarm on Pensions

Elizabeth Parisian of the American Association of University Professors, a leader of the Stop Funding Billionaires campaign, was also sharply critical. She called it “deplorable” that Anthropic was moving forward with its filing despite countless concerns raised by policymakers, financial stewards, unions, climate organizations and working people, adding that Anthropic employees had even resigned over concerns about the company.

Her main worry is workers’ money. Wall Street has rewritten financial rules to fast-track workers’ pensions and 401(k)s into A.I. I.P.O.s, as it did with SpaceX, she said. While pre-I.P.O. investors and insiders can cash out, working people’s retirement savings in index funds would be left exposed to future losses. Ms. Parisian compared the situation to the dot-com bubble and the 2008 financial crisis, and pointed to community opposition to new data centers and their climate impact.

The critics are getting support from New York: Mark Levine, the New York City comptroller, who advises the city’s pension funds with more than $325 billion in assets, recently sent a letter to Anthropic’s chief executive, Dario Amodei, urging him to delay the I.P.O. over safety concerns. Mr. Levine and Anthropic have since opened talks, but no delay has been agreed.

Anthropic’s Rationale

Anthropic explains the founders’ majority differently than its critics do. The seven co-founders, as a group, are “distinctly equipped to be stewards of our mission,” the prospectus says. In addition, a Long-Term Benefit Trust of five financially disinterested trustees with expertise in A.I. safety, national security and public policy is to elect a majority of the board. As a public benefit corporation, Anthropic is legally required to weigh the public interest alongside the interests of its shareholders. At the I.P.O., 5 percent of the shares go to a dedicated public benefit fund.

Anthropic is also active in politics: According to TechCrunch, the company backs Public First Action, a super PAC that supports A.I. legislation and in the summer backed the same candidate as the Guardrails Alliance.