Artificial Intelligence
Anthropic filing discloses fast growth and large losses
Anthropic’s S-1 showed rapid revenue growth, larger operating losses and extensive risk disclosures, while reporting and market speculation pointed to a high valuation and an IPO later this year.
- Anthropic filed an S-1 and warned investors, according to the Financial Times and Reuters, that advanced AI systems could lead to harmful behaviour including manipulation and blackmail.
- Reuters reported that revenue increased twelvefold in 2025 to almost $4.6 billion, while operating loss widened to $8.06 billion.
- The filing said Anthropic spent $7.33 billion on compute and infrastructure, and Reuters said roughly $34 billion of about $42 billion net loss was an accounting charge.
- The Financial Times said backers were discussing a valuation above $2 trillion, while Reuters said the offering was expected after the US midterm elections in November.
Anthropic filed an S-1 prospectus and, according to the Financial Times and Reuters, told investors that its AI systems could create serious risks. The filing also said more advanced models might behave in ways that included manipulation and blackmail.
Reuters said revenue rose twelvefold in 2025 to nearly $4.6 billion, but the operating loss increased to $8.06 billion. Anthropic spent $7.33 billion on compute and infrastructure, and Reuters reported that about $34 billion of roughly $42 billion net loss was an accounting entry tied to financing that could later convert into shares.
The filing said two customers accounted for almost a quarter of 2025 revenue, and that many large customers were not bound by long-term contracts. It also set out $518 billion of planned cloud, compute and infrastructure commitments over the coming years.
The Financial Times reported that backers thought a valuation above $2 trillion was possible. Reuters said that would follow a $965 billion valuation from May, exceed SpaceX’s roughly $1.8 trillion value at its June listing, and could come after the US midterm elections in November.
Named in this story
Companies
- Anthropic
- filed an S-1 and disclosed revenue, losses and risk factors
- OpenAI
- was cited as a rival that had confidentially filed for an IPO
- SpaceX
- was used as a valuation comparison
Organisations
- Financial Times
- reviewed the prospectus and reported on its contents
- Reuters
- reviewed the filing and reported the figures and timing
Governments and agencies
- US midterm elections
- formed the timing reference for the expected debut
How the source tells it
The piece used a trade-press register that mixed routine financial reporting with alarm over AI risk and promotional language about valuation.
- alarm loaded risk language in the filing was highlighted to stress possible harms from the technology
- hype valuation comparisons and industry-wide benchmark claims were used to magnify the offering
- speculation presented as fact forecast timing and price expectations were presented through cited sources and backers rather than confirmed events