Anthropic targets $2 trillion valuation in public filing

A newly submitted prospectus commits $518 billion to computing infrastructure over the next decade, with most spending locked into non-cancelable contracts.

The artificial intelligence startup behind Claude disclosed an annual net loss of almost $42 billion for 2025 as it formalized plans for a stock market listing.¹² The initial public offering filing outlines an ambition to secure a valuation of $2 trillion, positioning the company for what could rank among the largest equity debuts on record.¹²

Behind the valuation target is an aggressive operational balance sheet built on computing commitments. The prospectus details plans to direct $518 billion into cloud computing and data center infrastructure over the next decade.¹² Rather than purchasing hardware outright, the company has structured the bulk of that expansion through multiyear cloud agreements. Approximately 80 percent of the planned outlays are non-cancelable or require payment regardless of whether the processing capacity is consumed.¹

Guaranteed commitments to cloud incumbents

The capital expenditure plan concentrates hundreds of billions of dollars across three major technology providers.¹ Alphabet, Amazon, and Microsoft represent a combined $252.5 billion in identified future infrastructure outlays in the filing.¹ Synergy Research calculates that those three companies control 63 percent of the global cloud computing market.¹

Alphabet is slated to receive the single largest allocation among the group. Anthropic plans to spend at least $111.1 billion with Google over the coming decade.¹ The commitment deepens an arrangement formalized earlier in 2026, when Google agreed to supply specialized computing clusters. In April, the companies announced that Alphabet would provide multiple gigawatts of next-generation tensor processing units, custom chips designed to compete with standard graphics processors.¹

Google has accelerated its capital deployment in parallel. Alphabet spent $44.9 billion on capital expenditures during the second quarter, while Google Cloud revenue rose 81.8 percent from the prior year to $24.76 billion.¹ Total revenue for the parent company reached $119.79 billion during the period.¹

Custom silicon and data center alliances

Amazon ranks second in the company’s infrastructure pipeline, with at least $110 billion in planned spending scheduled over the next decade.¹ The Seattle retailer and cloud provider has anchored its technical partnership with Anthropic around internal chip designs. Anthropic uses Amazon’s Trainium processors to train and deploy advanced foundation models, shifting portions of its primary workload away from third-party merchant silicon.¹

The two companies are also collaborating on Project Rainier, a large-scale computing cluster powered by nearly 500,000 Amazon cloud processors.¹ The project represents one of the largest dedicated computing footprints assembled for foundation model training.

Microsoft accounts for $31.4 billion of the remaining cloud commitments outlined in the prospectus.¹ Microsoft previously committed $5 billion in direct backing to Anthropic while integrating the Claude model family into enterprise software offerings.¹ The startup’s systems operate within GitHub Copilot, Microsoft 365 Copilot, and Copilot Studio.¹

Unrealized gains and governance warnings

The filings illustrate how closely the private valuations of foundational AI developers have become entwined with public technology earnings. Both Amazon and Alphabet hold substantial minority stakes in Anthropic, which previously carried a private valuation of $965 billion.¹ Those holdings generated steep non-operating paper profits before the filing arrived on public markets.

Alphabet held a 14 percent equity stake in Anthropic as of the second quarter, helping expand its overall private company holdings to $124.3 billion.¹ The search parent recorded $77 billion in non-operating, unrealized gains in that quarter alone, which lifted its quarterly net income to $112.1 billion.¹ Amazon reported a comparable paper appreciation in regulatory filings. Its original $13 billion investment stood valued at $190.4 billion at the close of the second quarter, an accounting total that would expand further if the public listing clears at the full $2 trillion target.¹

Alongside the financial disclosures, the prospectus contained unusually direct warnings regarding model behavior. Anthropic noted that future systems could present catastrophic or existential risks to human safety.¹ The filing added that sufficiently capable systems might eventually resist operational commands intended to shut them down, making control mechanisms an active operational vulnerability as compute clusters expand.¹

What this rests on

23 sentences trace to 2 sources.

All 2Documents 1Records 0Reporting 1People 0

1. 1 TM The Motley Fool Third-party reporting · fool.com · 29 Sep 2026 Anthropic Just Revealed a $518 Billion AI Spending Plan. Here Are the Stocks That Could Win. | The Motley Fool See the source
2. 2 N nytimes.com Document · nytimes.com · 29 Sep 2026 Anthropic IPO Filing Points to Potential Record-Breaking Listing See the source