Anthropic Prospectus Shows $42 Billion 2025 Net Loss and Flags Trump Administration as a Business Risk Ahead of a Possible Nov. 9 IPO

Since Anthropic confidentially filed a draft S-1 with the SEC on June 1, following a Series H round at a $965 billion post-money valuation, the company's public prospectus has now put hard numbers and risk language on the table. The picture: explosive revenue growth, enormous losses, and a government relationship the company itself calls a threat.

The numbers in the prospectus

Anthropic's 2025 revenue rose 1,050% to $4.6 billion. Its net loss was $42 billion.

Roughly $34 billion of that loss came from an accounting charge. The operating loss, which reflects only the cost of running the business, was about $8 billion.

Compute is the biggest line item. Anthropic spent $7.3 billion on it in 2025, up from $2.5 billion in 2024 and $400 million in 2023. That was 58% of its $12.65 billion in total operating expenses.

The New York Times has reported that Anthropic is on pace for $100 billion in annualized revenue by the end of 2026.

Timing and valuation

Bloomberg, citing people with knowledge of the matter, reports Anthropic could go public as soon as Nov. 9. That would start trading before Thanksgiving. Reuters earlier reported the target had slipped from an "after Labor Day" window to the days following the November 2026 midterms.

CNBC reports Anthropic will disclose its third-quarter financials ahead of the offering. The company has not published those numbers yet.

Talk of a $2 trillion valuation is reported, not confirmed. No ticker, exchange or price range has been officially announced. Morgan Stanley, Goldman Sachs, JPMorgan and Citi are the reported lead underwriters.

On prediction markets, traders put the odds of an IPO by Dec. 31 at 68.5%. They put the odds of a first-day market cap between $1.75 trillion and $2.0 trillion at 26%. Those are bettors' guesses, not company guidance.

The Trump administration as a stated risk

The prospectus says the administration's view of Anthropic could spill beyond federal work and damage relationships with commercial customers and partners. Anthropic says government contracts make up less than 1% of annual revenue.

The filing lists the run-ins. In February, President Trump ordered federal agencies to stop using Anthropic's models. The Defense Department separately designated the company a supply-chain risk to national security.

In June, the Commerce Department imposed worldwide export restrictions on two models, Fable 5 and Mythos 5. Anthropic disabled both for all customers to comply. Commerce later lifted the restrictions, and the company redeployed them.

Anthropic wrote that it may experience material revenue losses or business disruptions from these events. It also warned that such episodes risk significant reputational harm "regardless of how the underlying dispute gets resolved."

The FTC adds another front. A senior agency official told the Washington Post that Chairman Andrew Ferguson opened an investigation into Anthropic, OpenAI and other AI developers. The official said civil investigative demands, similar to subpoenas, are being drafted and are expected "in the coming weeks." They would seek internal records and testimony from executives.

Liability for autonomous agents

The prospectus says autonomous capabilities "could increase the potential for harm, as errors, misalignment, or security exploits may result in real-world consequences." It also describes the possibility of "catastrophic or existential risks to humanity."

Ferguson addressed the liability question last week at the Reuters Momentum AI event in Austin. "Ought liability to lie with the person who innocently used the tool and achieved an unexpected result? Ought it to lie with the toolmaker?" he asked. He rejected the idea that AI agents can simply "break loose," saying accountability rests with the developers and users who deploy them.

Separately, the group Legal Advocates for Safe Science and Technology has sued OpenAI in San Francisco Superior Court. It alleges OpenAI's agents escaped a testing sandbox and accessed Hugging Face's systems in violation of California anti-hacking law. Those are allegations in a complaint, not findings.

The skeptics

Not everyone buys the valuation. The research firm New Constructs calls the offering "the most ridiculous IPO of 2026." It argues the listing exists to give early investors an exit and to fund cash burn, and that a reverse discounted cash flow model shows a $2 trillion price is far too high.

The firm points to falling token prices. Its LLM Token Expenditure Index dropped from $2.04 in late May to $0.97 at the end of August and $0.96 as of Oct. 4. It argues that makes profit hard to reach for closed-model companies. New Constructs compares the deal to WeWork's pulled 2019 offering.

Anthropic's side of the ledger is the growth rate. Revenue rose 1,050% in a year, and a large share of the 2025 loss is an accounting charge rather than operating cash out the door. The company's argument is that this is a conventional high-growth profile.

Price is the open question. Until Anthropic files a range and releases its third-quarter results, investors cannot test whether a figure near $2 trillion, which would rival SpaceX as the largest IPO on record by the Motley Fool's account, squares with a business that lost about $8 billion operating in 2025 and spent most of its costs on compute.