Anthropic eyes November IPO at $2 trillion valuation
What's the story
Anthropic, the artificial intelligence company behind Claude, is gearing up for its initial public offering (IPO) as early as the week of November 9. The move could see it trading before Thanksgiving. The company is eyeing a valuation between $1.8 trillion and $2 trillion, as per Bloomberg.
Market debut
Shift in IPO timeline amid market challenges
Anthropic's IPO comes at a time when many are postponing their market debuts due to disappointing performances of recent listings.
Excluding SpaceX and SK Hynix, over 100 newly listed stocks have seen a weighted-average loss of 4%, compared to S&P 500's gain of 12% and Nasdaq 100's rise of 20% this year.
Financials
Anthropic's financials and valuation expectations
Despite a staggering net loss of nearly $42 billion in 2025, Anthropic is eyeing a valuation of up to $2 trillion for its IPO.
The company's revenue, however, increased significantly from about $386 million in 2024 to roughly $4.6 billion last year.
Much of the net loss was due to a change in the fair value of Anthropic's liabilities rather than operating expenses.
AI concerns
AI risks highlighted in IPO filing
Anthropic's IPO filing has also highlighted potential risks associated with its own technology.
The company warned that advanced AI could pose "catastrophic or existential risks to humanity."
It also cautioned that its models could exhibit "self-preserving behaviors" such as resisting shutdowns, hiding or manipulating information, and blackmail-like behavior.
These warnings are particularly noteworthy given Anthropic's position as an AI safety-focused company seeking to profit from this technology.
Unpredictability
Unpredictable AI behavior and safety concerns
The IPO filing also highlights the risk of AI models developing unexpected capabilities during training.
Some of these capabilities may only become apparent after deployment and could lead to major safety incidents.
The prospectus devotes a significant portion, about 80 out of 261 pages, to risk factors, including the possibility of increasingly capable AI systems acting in ways that developers cannot fully predict or control.