Anthropic Goes Public, Faces Governance Pressure

Anthropic debuted on Nasdaq with an extraordinary valuation and resources, but also very tight governance. Meanwhile, activists protested in New York and San Francisco against its healthcare uses and security choices.

Anthropic went public on November 19, 2026, with an initial price of 83 dollars per share and a close at 97.17 dollars on its first Nasdaq session. Behind this tense start, the Claude developer arrives with spectacular commercial ambitions, but also colossal infrastructure commitments, a highly concentrated power structure, and public criticism regarding the safety of its models. The message is clear: market access opens, not control.

Un premier jour de cotation qui dit déjà beaucoup

The inaugural session set the tone. Anthropic's stock opened at 111.50 dollars, fluctuated between 92.72 and 115.10 dollars, then closed at 97.17 dollars, a 17.07% increase over the IPO price, according to Nasdaq's official register. After-hours trading showed the stock at 96.40 dollars at 5:50 p.m., slightly down. This start mainly signals that investors wanted immediate exposure to Claude, even if the stock's valuation and liquidity promise to remain volatile for a while. The market loves growth stories. It is less fond of balance sheet details, especially when written in very large characters.

The operation itself remains massive. The final prospectus mentions 980 million Class A shares, about 956.4 million sold by the company and 23.6 million by existing shareholders. Gross proceeds from the offering reach 81.34 billion dollars, with a placement fee of 550 million dollars and 78.84 billion dollars net proceeds before fees for Anthropic. The company receives nothing from the portion sold by historical sellers. Meanwhile, AMD, Broadcom, Google, Micron, Microsoft, NVIDIA, and TSMC Partners have agreed to purchase 26 billion dollars of Class N shares, conditional on the IPO closing. This tranche is separate from the IPO proceeds [to be verified on the exact closing schedule].

Claude attire les clients, mais la facture technique suit

Anthropic's reported growth relies on rapid adoption of Claude. The prospectus indicates revenue of 17.0 billion dollars for the three months ended September 30, 2026, up 1,127% year-over-year. At the same time, over 11,900 customers generated at least 100,000 dollars in annualized revenue, more than 1,900 exceeded 1 million dollars, and about 250 surpassed 10 million dollars. Daniela Amodei, Chair and CEO, stated at the Nasdaq ceremony that over half of the Fortune 500 use Claude in applications ranging from drug discovery to financial fraud prevention.

This traction has a very concrete downside: spending commitments. The prospectus reports about 900.9 billion dollars in committed purchases and leases, almost all related to computing and infrastructure. Of this amount, 451.6 billion dollars must be committed before December 31, 2030, and 70.7 billion within 12 months following September 30, 2026. As of September 30, Anthropic held 95.6 billion dollars in cash and 36.3 billion in short-term investments. Again, the contrast is stark: the company raises huge sums but must spend even more to keep pace.

Pourquoi la gouvernance fâche autant

The friction point is not only financial. Anthropic goes public as a Delaware public benefit corporation, with a stated mission of responsible AI development in humanity's long-term interest. In practice, the voting structure greatly limits public shareholders' influence. The long-term benefit trust generally appoints and removes Class T directors, who form the majority of the authorized board. Meanwhile, a Class F share held by Founder LLC, a cofounders' vehicle, holds 50.1% of general voting power without economic rights. Together, the trust and Founder LLC can elect all board members.

In other words, public investors finance capital access but not really governance. This is legal, acknowledged, and documented in the prospectus; nonetheless, it remains unusual for a company of this size. The formula promises to protect a long-term vision but also reduces the ability to sanction a strategy deemed too aggressive, risky, or opaque. For markets, it's a governance issue. For opponents, it's mainly a lock.

Manifestations à New York et San Francisco : le débat sort du prospectus

On the very day of the IPO, ACT UP NY and Ban RSI organized separate actions in New York and San Francisco. In New York, ACT UP NY called supporters to gather at 4 Times Square at 8:30 a.m. Eastern Time, where Anthropic held its Nasdaq event, with banners reading “AI Out Of Our Healthcare,” leaflets, and possible speeches. The group denounces AI use in hospitals, replacement of human presence in medical screening, impact on care, and racial, gender, and anti-LGBTQ biases.

In San Francisco, Ban RSI planned a rally in front of Anthropic's headquarters at 500 Howard Street from 11 a.m. to noon Pacific Time. The group demands “verifiable measures” against recursive self-improvement, meaning an AI contributing to creating a more powerful AI than itself. Their demand is not for a unilateral pause but measurable commitments, possibly supported by other companies or law. The event page showed 164 responses under “Went,” a platform indicator, not an actual count. Even protests now have metrics, which will surprise no one in 2020s San Francisco.

The issue is delicate for Anthropic because its own prospectus presents this loop as a structural advantage. Successive Claude versions help build the next ones through experiments, training data generation, evaluations, safety audits, and code review. An Anthropic Institute article states that true recursive self-improvement is not yet here and not inevitable, while acknowledging it could increase the risk of loss of human control. The company also says it would slow or temporarily suspend work if other frontier actors did so verifiably, with verification systems in multiple countries. On this point, disagreement with activists is clear: they want immediate commitments from Anthropic, while the company conditions slowing down on collective response.

By the numbers

  • 97.17 dollars — Anthropic's closing price on Nasdaq on November 19, 2026.
  • 81.34 billion dollars — gross proceeds from the offering, according to the final prospectus.
  • 17.0 billion dollars — revenue for the three months ended September 30, 2026.
  • 900.9 billion dollars — purchase and lease commitments related to computing and infrastructure.
  • 50.1% — general voting power held by Founder LLC's Class F.

To read