Dual-Class Stock at Anthropic, SpaceX Entrenches Founder

The artificial intelligence investment boom is being steered by a small circle of founders who have insulated themselves from shareholder oversight through dual-class share structures. Anthropic's planned public listing will grant its seven founders 50.1% of shareholder votes, cementing a governance model already deployed by SpaceX, Meta, and Google parent Alphabet. The result: investors pouring capital into AI must accept dramatically reduced influence over how their money gets deployed.
This governance architecture means five of the six leading US AI platforms operate under structures where ordinary shareholders lack the power to replace leadership or force strategic pivots. Of the major companies driving AI development, only Amazon, Microsoft, and Nvidia maintain conventional one-share-one-vote arrangements.
How Dual-Class Shares Concentrate Voting Power
In traditional corporate structures, each share carries one vote in shareholder decisions. Dual-class arrangements break this proportionality by granting certain shares multiple votes while others carry reduced or zero voting rights. Google's 2004 IPO popularized this model in Silicon Valley, giving founders Larry Page and Sergey Brin class B shares worth 10 votes each while public investors received single-vote class A shares.
Anthropic's structure follows this template while adding a unique oversight layer. CEO Dario Amodei and co-founders will control 50.1% of votes on most shareholder matters. Separately, an Anthropic Long-Term Benefit Trust whose members include former Federal Reserve Chair Ben Bernanke retains power to appoint and remove board members. The founders argue this arrangement prioritizes AI safety and national security over short-term profit pressures.
The Rise of Unequal Voting Structures
Dual-class shares emerged in 1898 with International Silver Co., but triggered backlash in the 1920s when Dodge Brothers went public with 1.7% equity holders maintaining full voting control. The New York Stock Exchange banned non-voting common stock from 1926 until 1986, when competitive pressure from Nasdaq forced rule changes.
The practice has accelerated dramatically in recent years. Last year, 35% of US IPOs and 64% of tech IPOs involved unequal voting rights, up from 12% and 36% respectively in 2016. Companies where founders retain material influence now represent roughly one-third of US stock market capitalization, more than double the share from a decade ago.
Officially Launched on November 16, 2026
Anthropic's governance plan was officially released on November 16, 2026, as part of its pre-IPO disclosures. The announcement comes as the company prepares for a public listing expected to value it among the largest AI developers globally. OpenAI's governance remains unclear, with CEO Sam Altman holding no direct equity stake and control resting with a nonprofit foundation, though its planned listing could add another founder-insulated AI giant to public markets.
Investor Concerns and Performance Trade-Offs
Critics argue dual-class structures undermine shareholder democracy precisely when it matters most. New York City Comptroller Mark Levine, overseeing city pension funds, calls Anthropic's plan part of a "very worrisome trend" shifting power from shareholders to management. The concern intensifies given AI companies' massive capital requirements and uncertain revenue projections.
Proponents counter that founder control enables long-term thinking over quarterly pressures. An index tracking US founder-led companies has outperformed the S&P 500 by more than 5 percentage points annually since 2016. However, research shows dual-class valuation premiums typically dissipate as companies mature and founders' voting power increasingly exceeds their financial stakes. Organizations like the Council of Institutional Investors advocate for seven-year sunsets on multi-class structures.
The Indexing Dilemma
The rise of passive investing has shifted bargaining dynamics around governance. Active fund managers can refuse to buy poorly-governed companies, creating pressure for change. But once a company enters a major index, passive funds tracking that benchmark must hold it regardless of governance structure. S&P Dow Jones Indices briefly banned multi-class companies from benchmarks like the S&P 500 after industry consultation, then reversed course in 2023.
This leaves institutional investors with limited options. Pension funds, sovereign wealth funds, and other large investors collectively own 53% of publicly-traded SpaceX shares despite its governance structure. After SpaceX's June 2026 listing that helped make Elon Musk the world's first trillionaire, shares soared to a price-to-sales ratio of 33, five times the Nasdaq 100 average, illustrating how founder cult followings can translate into financial leverage.
What This Means
Anthropic's governance plan crystallizes a fundamental tension in AI investment: capital is abundant, exposure opportunities are scarce, and founders hold the leverage to demand money without ceding control. As the industry scales from billions to hundreds of billions in capital deployment, the structural inability of shareholders to force course corrections if optimism proves overdone represents a significant governance experiment. Whether founder vision or shareholder accountability better serves AI development remains an open question, but investors seeking exposure to the technology's potential increasingly have no choice but to accept the former.
on Emergent today
.avif)




