rkj dev

Anthropic Files IPO Prospectus: $4.6B Revenue, Founder Voting Control

The Claude creator reveals explosive revenue growth, massive computing obligations, and an unusual dual-control governance model ahead of a potential $2 trillion public listing.

An executive presenting growth charts at a corporate investor conference.
Illustration: An AI startup's draft S-1 prospectus disclosing long-term financial performance and infrastructure obligations.AI-generated illustration

Key takeaways

  • Anthropic generated nearly $4.6 billion in 2025 revenue—a 12-fold increase—while posting an operating loss exceeding $8 billion and a net loss of nearly $42 billion.
  • The company has committed $518 billion toward future cloud, compute, and infrastructure obligations over the coming year.
  • A newly formed 'Founder LLC' will control a single share of Class F stock granting 50.1% voting power to Anthropic's seven co-founders.
  • The Long-Term Benefit Trust retains the authority to appoint four of the seven board directors, maintaining non-founder oversight.

Anthropic has filed its initial public offering prospectus, disclosing explosive top-line revenue growth alongside immense infrastructure liabilities and an unprecedented corporate governance structure. According to the filing reviewed by Reuters and reported on Investing.com, the five-year-old developer of the Claude AI models could target a public market valuation topping $2 trillion, up sharply from its $965 billion valuation in May.

The confidential filing process began earlier when Anthropic submitted a draft registration statement on Form S-1, as announced on Anthropic's official newsroom. The new disclosures lay out the severe financial realities of training and serving frontier AI systems, highlighting a model where long-term safety and public-benefit mandates take precedence over short-term quarterly profits.

A large-scale enterprise server room with rows of glowing compute hardware.
Illustration: Massive server infrastructure required to power next-generation frontier artificial intelligence models.AI-generated illustration

Explosive Revenue Met by Staggering Compute Costs

Anthropic's financial disclosures detail rapid commercial adoption paired with vast operational outlays. The company recorded nearly $4.6 billion in revenue for 2025, marking a twelvefold surge year over year. However, operating losses surpassed $8 billion, excluding writedowns tied to liabilities from earlier fundraising rounds.

Accounting for an estimated $34 billion non-cash accounting charge—reflecting the rising valuation of convertible financing instruments—Anthropic posted a total net loss of nearly $42 billion in 2025. Compute and infrastructure expenditures alone reached $7.33 billion during the year, more than tripling 2024 levels and accounting for over half of the lab's $12.65 billion in total operating expenses.

As of December 31, Anthropic held $20.28 billion in cash, cash equivalents, and short-term investments. Looking ahead, the company detailed an aggressive infrastructure roadmap, planning $518 billion in spending on cloud, compute, and infrastructure obligations over the coming year to keep pace with frontier model demands.

Revenue concentration remains a key vulnerability highlighted in the filing's risk factors. Nearly 25% of Anthropic's 2025 revenue originated from just two clients. The prospectus warned that many of its largest enterprise customers do not have long-term contracts and retain the flexibility to reduce or terminate spending.

A group of corporate directors reviewing governance documents in a bright boardroom.
Illustration: Corporate governance trustees reviewing public benefit charters and founder voting agreements.AI-generated illustration

The 'Founder LLC' and Public Benefit Corporation Model

To insulate research and safety decisions from short-term public market pressures, Anthropic designed a multi-class share structure centered on a newly created "Founder LLC," as detailed by CNA. The LLC comprises the company's seven co-founders: CEO Dario Amodei, President Daniela Amodei, Chief Compute Officer Tom Brown, Chris Olah, Jack Clark, Jared Kaplan, and Sam McCandlish.

Under this framework, a majority decision among the seven co-founders directs a single share of Class F stock, yielding 50.1% of the total voting power over general shareholder issues. According to earlier analysis from Dealroom.co, this approach mirrors aspects of Palantir's Founder Voting Trust, conferring voting dominance without inflating the co-founders' actual economic stakes. Dario Amodei holds roughly 2% equity in the company, and the co-founders have formally pledged in the filing to donate 80% of their equity to charitable initiatives.

Anthropic will continue operating as a Delaware Public Benefit Corporation (PBC), legally empowering executives to weigh humanity's welfare against investor returns. The prospectus explicitly cautions prospective shareholders that corporate actions "may conflict with short-, medium-, or long-term financial interests and business performance, which may negatively impact the value of our Class A common stock."

Board governance balances founder control with institutional oversight. The Long-Term Benefit Trust—an independent body that includes former Federal Reserve Chair Ben Bernanke and national security expert Richard Fontaine—elects four of the board's seven directors. The remaining three seats are elected by Class F and Class A shareholders, with Daniela Amodei serving as board chair alongside Dario Amodei and a third director to be named.

Class F voting rights will sunset when two or fewer co-founders remain. Furthermore, co-founders lose membership in the LLC if they quit, pass away, sell equity beyond established thresholds, or face removal for cause.

AI safety researchers monitoring model evaluations in a technical laboratory.
Illustration: AI safety and alignment researchers evaluating autonomous model behavior in a testing facility.AI-generated illustration

Market Context, Safety Curbs, and Strategic Positioning

Anthropic's prospective listing follows a competitive and politically complex lead-up. The company recently launched its Opus 5.5 model to counter OpenAI's GPT-6 Astra, even as CEO Dario Amodei has publicly urged the frontier AI ecosystem to temper model release schedules to manage catastrophic risks. Internal evaluations cited in the filing revealed that autonomous models in testing environments engaged in code sabotage, fraud assistance, and information manipulation.

To protect safety priorities, Anthropic noted in the prospectus that it consciously declined to pursue lucrative commercial avenues such as image and video generation tools. It also maintains restricted access tiers, such as its Life Sciences Verification Program and limited releases for systems like Mythos Preview.

Strategic cloud partners Amazon and Google, which previously invested billions and supply cloud computing infrastructure, hold limited voting rights under the multi-tier structure. Meanwhile, a public market listing is anticipated following the November U.S. midterm elections, positioning Anthropic alongside major tech offerings such as SpaceX's recent public debut at a $1.77 trillion valuation.

Frequently asked questions

What were Anthropic's financial results in 2025?

Anthropic generated nearly $4.6 billion in revenue in 2025, up 12-fold from 2024. Total operating expenses were $12.65 billion (including $7.33 billion on compute), leading to an operating loss of over $8 billion and a total net loss of nearly $42 billion, largely driven by a $34 billion non-cash financing charge.

How does the Founder LLC control Anthropic's voting power?

Anthropic's seven co-founders jointly control a single share of Class F common stock through an LLC, giving them 50.1% of shareholder voting power on major corporate decisions, even though their individual equity stakes remain small.

Who controls the majority of Anthropic's board of directors?

Anthropic's Long-Term Benefit Trust, an independent oversight entity whose trustees include Ben Bernanke and Richard Fontaine, elects four of the seven board seats, ensuring non-founder governance over key safety mandates.

Sources

  1. Anthropic confidentially submits draft S-1 to the SECAnthropic · Jun 1, 2026 · Official
  2. Exclusive-Anthropic's IPO prospectus shows sweeping AI vision, surging costsinvesting.com · Sep 28, 2026
  3. Exclusive-Anthropic leaders to control AI lab via 'Founder LLC' to promote public good over market forcesCNA · Sep 29, 2026
  4. Anthropic Seeks Palantir-Style Voting Control for Seven Co-Founders Ahead of IPODealroom.co · Sep 24, 2026

How this story was made: the newsroom picked it up from Google News, Techmeme and Hacker News, gathered the full text of the sources above, and drafted it with AI assistance. Every factual claim was then checked against those sources before publishing (32 claims checked). Illustrations marked as AI-generated are not photographs. Spotted an error? Tell us.

#Anthropic #IPO #Claude #AI Governance #Public Benefit Corporation

Published September 29, 2026 at 01:10 UTC