Anthropic, the developer of Claude, is planning one of the biggest IPOs in history as a Delaware public benefit corporation, while its draft prospectus warns that increasingly autonomous AI models could evade oversight, resist shutdown, manipulate operators and pose catastrophic or existential risks to humanity. The company reported $4.59 billion of 2025 revenue, a $41.97 billion GAAP net loss and an $8.06 billion operating loss; roughly $34 billion of the net loss reflected a non-cash charge tied to convertible financing. Amazon and Google each accounted for about 12% of 2025 revenue and also hold approximately 21% and 15% of Anthropic, respectively. Anthropic committed about $518 billion to cloud capacity, chips and related infrastructure, including more than $100 billion of planned spending with Amazon Web Services over the next decade. It held $20.28 billion in cash, cash equivalents and short-term investments at the end of 2025, reported second-quarter 2026 revenue above $11.5 billion, and said its annualized revenue run rate rose to $47 billion by May 2026 and exceeded $65 billion by July. Its amended certificate of incorporation says its public benefit is to responsibly develop and maintain advanced AI for humanity’s long-term benefit, requiring directors to balance investor returns, that mission and affected stakeholders. Reuters reported that an offering expected after the November 3 U.S. midterm elections could seek a valuation above $2 trillion, subject to market conditions and other factors.