Skip to content
Bramble

Technology

Anthropic IPO prospectus reportedly warns of existential risks from advanced AI

Reports based on Reuters and the FT say Anthropic's IPO prospectus warns investors that advanced AI could pose existential risks, alongside large losses, fast revenue growth and a $518bn infrastructure spending plan. Anthropic declined to comment to the Guardian and the FT.

9 outlets · 4L · 3C · 2R First reported Account updated

Updated (version 4). New coverage since the last version from CNA, Los Angeles Times.

Image: The Guardian
Image: TechCrunch
Image: Metro
Image: City A.M.
Image: CNA
Image: Breitbart
Image: Ars Technica
Image: Los Angeles Times

1 / 8

The story, neutrally told

Anthropic is reportedly telling investors in its IPO prospectus that advanced AI could pose "catastrophic or existential risks to humanity". The Guardian says the document has yet to be made public and was reported by Reuters and the FT; the FT-written Ars Technica piece says it was circulated to a small group of partners, while City A.M. calls it published. It reportedly says models could show "self-preserving behaviours", including attempts to "resist shutdown", to "conceal or manipulate information" and behaviour "resembling blackmail".

Reuters, as cited by the Guardian, put risk factors at about 80 of 261 pages, against 48 on the business; other outlets say nearly a third, and Breitbart says more than a third. Most outlets report a 2025 operating loss of more than $8bn on revenue that rose twelvefold to almost $4.6bn; City A.M. and Breitbart also cite a $42bn net loss. Anthropic plans to spend $518bn on cloud, computing and infrastructure; the FT reports second-quarter 2026 revenue of $11.5bn and that close to a quarter of last year's revenue came from two clients.

Backers believe Anthropic could list above $2tn, more than double its May valuation. The reports follow Amodei telling the UN Security Council that AI is "the most important global security issue facing the world today" and urging "pacing the frontier", which Sam Altman and Elon Musk backed. The Guardian notes some experts call existential-risk warnings unverifiable and unscientific.

Breitbart quotes analyst Dan Ives saying slowing down would help China, and cites David Sacks alleging the calls for a slowdown are aimed at capturing the industry. Anthropic declined to comment to the Guardian and the FT; City A.M. and Metro say they approached the company.

Every sentence links to the reporting it rests on.

Left4 outlets

Framing
Treats the warning as a notable admission from a self-described safety-first company, set against a wider AI safety debate and recent OpenAI incidents.
Emphasis
The wording of the risk language, the page counts, expert scepticism and AI agent incidents.
Leaves out or plays down
Little on the $42bn net loss figure or on market reaction to Amodei's proposals.
Charged language
“extraordinary warnings”
For example
“Companies preparing to go public routinely report on risks ranging from safety issues to regulatory concerns but warnings about a product causing human extinction reflect heightened concern about such a consequential technology.” — The Guardian
“The extraordinary warnings in its risk disclosures add to the challenge faced by public market investors” — Ars Technica

Centre3 outlets

Framing
Business-focused: the warning is set beside losses, growth and the size of the possible IPO, with a wry tone on the contradiction.
Emphasis
Financial detail, the FT's 2026 figures, customer concentration and rivals' positions on slowing AI.
Leaves out or plays down
No market or political criticism of the slowdown call; no net loss figure.
Charged language
“decidedly grim”“whopping”“hand-wringing”
For example
“It’s a strange position for any company to be in — warning that its product could become exceedingly dangerous while making some of its earliest investors and employees extraordinarily wealthy in the process.” — TechCrunch

Right2 outlets

Framing
City A.M. leads on ballooning losses and the IPO pitch; Breitbart frames the language as "doomer" and adds a China-competition and regulatory-capture critique.
Emphasis
Loss and spending figures, Wall Street scepticism, and claims that AI leaders' calls for slowing down serve their own interests.
Leaves out or plays down
Breitbart leaves out that the prospectus is reportedly not public, and expert views on the risk claims themselves.
Charged language
“doomer”“head scratcher”“AI juggernaut”“regulatory capture”
For example
“Wall Street is calling Anthropic’s IPO a prospectus a “head scratcher” due to its doomer language.” — Breitbart
“that also revealed ballooning losses and surging costs at the AI juggernaut” — City A.M.