Story perspectives

Anthropic IPO Prospectus Warns Models Could Pose Existential Risks

By Drooid · Sep 29, 2026, 6:02 AM UTC · How we work

Summary

Anthropic Files IPO

Anthropic filed its IPO prospectus warning its models could pose catastrophic or existential risks.
The filing creates a Founder LLC giving seven co-founders 50.1% voting control via Class F share.
Anthropic posted a $42 billion net loss for 2025 while planning $518 billion cloud spending.
Revenue rose twelve-fold to $4.6 billion in 2025 while operating loss topped $8 billion.
CEO Dario Amodei earned nearly $18 million in 2025, mainly from stock awards.

Summary

Anthropic Prioritizes Safety

Anthropic allocated roughly 6% of July compute power to safety research, per its prospectus.
The filing says safety work is resource-intensive, competing with compute capacity and high-cost AI talent.
CEO Dario Amodei published a 4,000-word essay urging the industry to pace frontier AI development.
Anthropic pledged to publicly share more data on how current models train future generations.

History

Historical Context

OpenAI announced GPT-2 in February 2019 and limited its code, citing misuse risks; the 1.5-billion-parameter model spurred a voluntary research pause and a U.S. AI policy debate. That caution shaped later competitors, and Anthropic’s 2025 IPO prospectus emphasizes safety, echoing the global industry regulatory governance framework first set by OpenAI’s 2019 restraint.

1 of 3

Anthropic Files IPO

Anthropic filed its IPO prospectus warning its models could pose catastrophic or existential risks.
The filing creates a Founder LLC giving seven co-founders 50.1% voting control via Class F share.
Anthropic posted a $42 billion net loss for 2025 while planning $518 billion cloud spending.
Revenue rose twelve-fold to $4.6 billion in 2025 while operating loss topped $8 billion.
CEO Dario Amodei earned nearly $18 million in 2025, mainly from stock awards.

1 / 3