Anthropic's IPO: VCs Bank Billions, But Amazon's Looming Influence Sparks Leadership Questions

Founded in 2021 with the ambitious mission to ensure humanity can control and benefit from artificial intelligence in the long term, Anthropic is now approaching its initial public offering (IPO). However, its corporate governance structure is meticulously designed to prioritize this AI safety mission, creating a unique framework that significantly limits the influence of common shareholders over major decisions, including board appointments.

Anthropic operates as a public benefit company (PBC), a for-profit corporation legally empowered to balance a public mission alongside traditional profit motives. This status is crucial, as it protects the board's prerogative to make decisions concerning AI safety without being solely driven by shareholder value, thereby insulating it from potential lawsuits by dissenting investors. The current board is chaired by Anthropic president Daniela Amodei and includes CEO Dario Amodei, Netflix cofounder Reed Hastings, former Microsoft and GM CFO Christopher Liddell, Novartis CEO Vas Narasimhan, and Spark Capital partner Yasmin Razavi. There is currently one vacancy on the board.

A cornerstone of Anthropic's governance is the Long-Term Benefit Trust (LTBT), an independent panel explicitly tasked with maintaining the company's focus on its mission to benefit humanity. This trust holds a special class of stock that grants it the power to elect and remove a majority of the company’s board members, known as Class T directors. The LTBT is chaired by global health expert Neil Shah, formerly of GiveWell, and its members include former Federal Reserve chair Ben Bernanke, national security expert Richard Fontaine, and Harvard physicist Christopher Stubbs. Like the main board, the LTBT currently has one vacancy, which its existing members will fill.

Further concentrating control, Anthropic's seven founders jointly possess a special class of stock that gives a majority of them 50.1% of the total shareholder voting power. This enables them to elect three members to the board of directors. Should the founders fail to reach a majority agreement for any reason, their shares would then be voted in proportion to holders of common stock. This founder control is not perpetual; it is set to expire ten years after the IPO, or automatically terminate when fewer than three of the original founders remain with the company.

Despite this intricate web of control, certain checks and balances exist. The board's nominating committee has the authority to decline the LTBT's nomination of a “Specified” director or the removal of such a director without cause (i.e., without allegations of misconduct). In such a scenario, the decision is then put to a vote among the company's employees. This mechanism, similar to the employee power seen during OpenAI's governance crisis, suggests that employee loyalty could significantly influence outcomes. Common shareholders, however, will only truly gain control of the company if and when neither the founders' special class of stock nor the LTBT exists any longer, a condition that depends on the eventual sunsetting of founder control and the LTBT either being voted out by 85% of common shareholders or stepping down with board approval.

The financial prospects surrounding Anthropic's IPO are significant, particularly for its strategic partners. While many venture capital (VC) firms, including Menlo Ventures, Sequoia, Sound Ventures, and Spark Capital, are poised for substantial returns, it's notable that none of these firms hold a stake of 5% or more, the typical threshold for inclusion in a principal stockholders' table. This unusual situation stems from Anthropic's exceptionally high private valuation, estimated at $965 billion. At this scale, even a 1% stake is worth nearly $10 billion, meaning large investments translate to relatively small percentage ownerships.

The biggest financial winners from Anthropic’s IPO are largely its strategic partners. Amazon, in particular, is positioned as a primary beneficiary. Its involvement began in 2023 with a commitment to invest up to $4 billion. Since then, Amazon has invested billions more, primarily through convertible notes. As of June 30, Amazon held convertible notes with $6.2 billion in principal outstanding, representing Anthropic’s entire debt load. As Anthropic's valuation dramatically increased across successive funding rounds, the shares into which these notes will convert became far more valuable. Amazon's own filings valued its note holdings at an astonishing $97.9 billion as of June 30. This surge in value necessitated Anthropic to record a $72.4 billion non-cash charge on its income statement for the first half of 2026, categorized as “change in fair value of convertible notes, derivative liabilities, and SAFEs.” The convertible notes could convert into as many as 1.84 billion shares, though Amazon is capped at 12% of Anthropic’s “partially diluted” share count, a limit it can waive. Beyond the notes, Amazon further invested $10 billion in nonvoting preferred stock in the second quarter, bringing its total investment to approximately $18 billion ($8 billion in notes from 2023-2025 and $10 billion in preferred stock). In addition to these equity gains, Amazon stands to earn significant revenue from Anthropic's commitment to spend at least $110 billion on AWS over the next decade.

Google is another strategic partner set for a major windfall. Early last year, it held about 14% of Anthropic, according to court filings, and has since invested more, including an additional $10 billion in April. Although subsequent large funding rounds likely diluted Google's stake to an estimated 5.6%, it has also secured at least $114 billion in cloud computing commitments from Anthropic through 2033. The S-1 filing itself highlights a potential conflict of interest, warning that some of these cloud providers also develop rival AI models, indicating that their interests may not always align perfectly with Anthropic’s mission.