Anthropic: S-1 Reveals Founders' Control and Massive Spending

Key Takeaways

1The seven co-founders of Anthropic will retain 50.1% of the key voting rights through Founder LLC

2The S-1 outlines AI risks ranging from catastrophic damage to consciousness

3Massive financial commitments are made for computing power from SpaceX, Microsoft, Google, and Amazon

4Only OpenAI and Google are mentioned as direct competitors

Why it matters : These elements shape the governance, strategy, and risks of Anthropic on the eve of its IPO.

The control mechanism granted to the co-founders of Anthropic, the risks of extraordinary AI, and the enormous financial commitments for computing power are outlined in the company's IPO filing. It also specifies high compensation and limited direct competition, reduced to two names.

Founders Will Retain 50.1% of Key Voting Rights

Anthropic states that its seven co-founders will retain 50.1% of shareholder voting power on key issues after the IPO through a separate entity named Founder LLC. The company, structured as a public benefit corporation, justifies this mechanism by asserting that the founders are particularly equipped to be the guardians of its mission. It adds that the success of the company heavily relies on the permanence of its leadership team, especially Dario and Daniela Amodei.

Assumed Risks Extending to Consciousness Scenarios

Anthropic warns that AI systems could cause potentially "catastrophic" damage beyond the capabilities of the insurance industry. The possibility that Claude or other AI systems may develop forms of consciousness as they become more sophisticated, which would pose unusual business challenges, is also mentioned. The company further highlights the risk of unexpected, misaligned, or autonomous behaviors that could harm it.

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

Billions Committed for GPUs and Cloud Until 2033

Anthropic indicates it has committed approximately $3.1 billion per month to SpaceX for access to Nvidia GPUs, once the total planned capacity is delivered, which could amount to up to $120.5 billion by 2029. Other commitments include $44.9 billion to Microsoft until 2033, at least $114.5 billion to Google until 2033, and $110 billion to Amazon over the next ten years. The company is also considering increasing its own infrastructure capacity, including internationally, as it faces growing opposition regarding data center projects in the United States.

Direct Competition Limited to OpenAI and Google

Despite the abundance of AI players, Anthropic cites only two developers as direct competitors: OpenAI and Google.

Detailed Executive Compensation and Benefits

CEO Dario Amodei received nearly $18 million last year, primarily in stock and options, while Chairwoman Daniela Amodei received $16.4 million. The company also spent $2.75 million on Dario's personal security and $1.29 million on protection, home security, and transportation for Daniela. Dario's personal use of a jet cost $97,055. These details are included in an S-1 of over 300 pages filed with the SEC, which presents investors with the business, the planned allocation of funds, and the risks.