Anthropic Before the IPO: Trust and Founders Retain Control

Key Takeaways

1An independent internal trust can elect and revoke the majority of the board of directors

2The founders hold a special class of shares giving them 50.1% of the voting rights and the ability to elect three directors

3Employees may be called to vote on certain arbitrations regarding board seats

4Clauses provide for the end of the founders' special control after ten years or in the event of departures, with a possible return of control to ordinary shareholders

Why it matters : This atypical governance limits the power of future ordinary shareholders and formalizes unprecedented internal checks and balances in tech.

The AI company founded in 2021 is moving towards an IPO with a governance structure that largely eludes future ordinary shareholders. An internal trust, a class of founder shares, and employee oversight on certain choices create a unique balance, accompanied by clauses for an eventual exit.

Locks and Possible Exits Frame Special Power

The board of directors of Anthropic can oppose certain decisions made by the Long-Term Benefit Trust (LTBT) regarding the appointment or dismissal without cause of a "Specified" among its four directors, by submitting the matter to an employee vote. This mechanism allows employees to intervene in the arbitration of these situations if the nomination committee seeks their opinion. If the founders fail to achieve a majority, the voting power of their special class of shares is then distributed proportionally with the holders of ordinary shares, which can increase the influence of the latter. Furthermore, clauses provide for the end of the founders' control: ten years after the IPO, they can choose to relinquish it, or this control automatically ends if fewer than three founders remain. After this deadline, the LTBT remains in place unless 85% of ordinary shareholders vote to terminate it, or if the trust decides to withdraw with the board's consent. If no special class of shareholders remains, control then reverts to the ordinary shareholders.

Today, a Trust Can Choose the Majority of the Board

Anthropic has established an independent Long-Term Benefit Trust (LTBT), tasked with ensuring that the company remains focused on its mission of benefiting humanity. This trust holds a class of shares that allows it to elect and revoke the majority of the board members, referred to as Class T directors. The LTBT is chaired by Neil Shah, a global health expert and former CEO of GiveWell, and includes members such as Ben Bernanke, Richard Fontaine, and Christopher Stubbs. There is currently a vacancy within the trust, which its members will fill themselves. The board retains oversight over certain LTBT decisions related to board composition, through the established mechanism.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

The Founders Hold 50.1% of Special Voting Rights

The seven founders of Anthropic collectively hold a special class of shares that grants them, by majority, 50.1% of the voting rights of shareholders. This position allows them to elect three members of the board of directors. If no majority emerges among them, their voting rights are then distributed proportionally with those of the ordinary shareholders.

A Public Benefit Board, Protected to Oversee AI Safety

Anthropic has the status of a public benefit corporation, with the goal of promoting AI safety. This legal framework allows the board to prioritize safety-guided decisions, without being limited solely to the pursuit of shareholder value, and reduces the risk of investor litigation. The board is chaired by Daniela Amodei and includes Dario Amodei, Reed Hastings, Christopher Liddell, Vas Narasimhan, and Yasmin Razavi. There is currently one vacancy on the board.

The OpenAI Precedent Illustrates the Unpredictability of Balances

At OpenAI, the board of directors briefly sidelined Sam Altman for reasons related to a lack of transparency, before reinstating him and implementing a new governance structure. This episode shows that complex structures can lead to unexpected situations. In the case of Anthropic, the combination of a trust with extensive rights and a special class of shares for the founders limits the influence of ordinary shareholders on major decisions, particularly regarding board composition.