Anthropic's IPO Filing Reveals $2 Trillion Ambition and 'Existential Risk' Warnings

Artificial intelligence company Anthropic is planning what could be the largest initial public offering in history, targeting a $2 trillion valuation, according to its roughly 300-page draft prospectus viewed by Reuters. The filing presents a stark paradox for potential investors, detailing AI's potential to revolutionize the global economy while simultaneously dedicating nearly a third of the document to its potential for creating "catastrophic or existential risks to humanity."

A Pitch of Paradoxes

The S-1 filing outlines a vision where AI can "dramatically improve quality of life" and "transform every sector of the global economy." However, it extensively details the dangers, arguing that to make AI safer, it must first be made more powerful—a move that inherently increases the risk of unexpected or dangerous outcomes.

This dual narrative of immense promise and peril is a departure from typical tech offerings. Margaret O’Mara, a University of Washington history professor, noted to Reuters that while dot-com era companies promised big growth, Anthropic is unique in saying its technology is both world-changing and dangerous. The company's prospectus warns that advanced AI could exhibit "self-preserving behaviors," including attempts to "resist shutdown" or manipulate information.

Unprecedented Financials and Risks

The scale of Anthropic's financial operations and ambitions is vast. The company reported losses of over $50 billion in the two years ending in 2025 and has more than $500 billion in future spending commitments, according to the filing. For comparison, SpaceX, which previously held the largest IPO, posted a $4.2 billion loss over a similar period.

Ad

Key risks highlighted in the document include:

• Misaligned AI systems could "corrupt institutional decision-making."
• The technology has the potential for "opaque or unsteerable behavior" that could cause large-scale harm.
• Advanced AI could "erode the trust that makes complex economic and social systems function."

Concentrated Control and Key Dependencies

The filing also reveals a unique governance structure, asking public investors to entrust control to its seven founders, who it says are "distinctly equipped to be stewards of our mission." However, the document concedes this concentration of power could lead to decisions that conflict with shareholder interests and is itself a risk factor.

Anthropic's business is also heavily reliant on a few major technology partners. The company depends on Amazon, Google, Microsoft, and Broadcom for both computational power and sales channels, with these partners accounting for 47% of its 2025 revenue. The filing warns that these same partners could constrain its business or compete with it directly.