Full Breakdown
Anthropic’s $80 B Half-Year Loss Raises Questions Ahead of Planned $2 T IPO
By Drooid · · How we work
Core Event: Massive Loss and IPO Plans
Anthropic reported a net loss of more than $80 billion for the first half of the year, up sharply from a $13 billion loss a year earlier. The loss was driven primarily by $72 billion in accounting charges tied to changes in the value of convertible notes and other fundraising instruments. Revenue rose to $16.3 billion in the six months ended June 30, up from $1.1 billion a year prior. The San Francisco-based AI firm is preparing an IPO slated for November, targeting a valuation of up to $2 trillion.
Background & Context: AI-Race Spending and Governance
Anthropic’s model depends on massive compute capacity. The company has committed to future cloud-computing contracts estimated at $450 billion over the next seven to ten years and anticipates total spending on additional resources of up to $750 billion. Infrastructure expenditures jumped to nearly $9 billion in the first half, more than tripling the prior year’s outlay.
The draft prospectus shows a governance structure that would give the seven co-founders 50.1 % of voting power through Founder LLC, potentially limiting shareholder influence over safety decisions. Anthropic pledged 80 % of the founders’ personal shares to philanthropy and earmarked 5 % of total shares for a public-benefit fund.
Data & Statistics
- Net loss (first half): $80.7 billion
- Accounting charges: $72 billion
- Revenue (first half): $16.3 billion
- AI infrastructure spend (first half): $9 billion
- Future cloud-computing contracts: $450 billion
- Potential total compute spend: $750 billion
- Two largest customers: roughly 20 % of revenue
- Valuation expectations: $1.5 trillion (investors) vs. $2 trillion target
Official Statements & Responses
The company warned that AI models might exhibit “self-preservation” traits or generate “blackmail-like” responses that could hinder shutdown attempts. In response, Anthropic announced expanded independent reviews, granting six researchers from the nonprofit Machine-Intelligence-Research-Institute (METR) access comparable to internal safety teams. It is also lobbying Congress for legislation that would empower the Department of Commerce and federal courts to block AI models deemed dangerous.
Conflicting Reports & Gaps
Valuation expectations diverge: some sources cite a $2 trillion target, while investors referenced at the meeting favor $1.5 trillion. Anthropic has not publicly commented on the debate, and Reuters was unable to verify certain risk disclosures.
Verbatim Quotes
- “Risk the IPO? I don’t think so,” — Lior Prosor
- “This is not going to be straight up and to the right forever,” — Lior Prosor
What’s Next
Anthropic is expected to file its formal IPO documents within weeks. The company will continue negotiating compute-capacity contracts and pursuing legislative proposals that could give U.S. regulators authority to block high-risk AI models. The outcome will shape IPO pricing and market perception of AI-sector risk.