A Sneak Peek Into Anthropic’s Business

Summary of A Sneak Peek Into Anthropic’s Business

by The Wall Street Journal & Spotify Studios

18m•October 21, 2026
Listen to Episode

Overview of A Sneak Peek Into Anthropic’s Business

This episode of The Journal examines Anthropic’s draft S-1 ahead of a potentially blockbuster IPO, expected as early as November. Using the confidential filing, the hosts break down how the AI company behind Claude is growing at extraordinary speed — but also facing huge risks, including customer concentration, massive compute costs, and uncertainty around AI safety and long-term demand. The central question: is Anthropic shaping up to be one of the most valuable companies in the world, or is the AI boom showing signs of strain?

Key Financial Takeaways

  • Explosive revenue growth

    • Anthropic reportedly brought in about $16 billion in revenue in the first half of 2026, up from around $1 billion in the same period a year earlier.
    • The company also indicated third-quarter 2026 revenue was around $16 billion, suggesting continued rapid growth.

  • Huge valuation expectations

    • The company’s IPO could reportedly value Anthropic at up to $2 trillion and raise around $100 billion, making it one of the most anticipated IPOs ever.

  • Massive reported loss

    • Anthropic posted an $80.7 billion net loss in the first six months of 2026.
    • Most of that — about $72 billion — was an accounting adjustment, tied to the value of shares previously promised in exchange for loans, rather than cash actually spent.

Customer Concentration and Revenue Risk

  • Anthropic disclosed that 20% of first-half revenue came from just two customers.
  • The company expects a small number of large customers to account for most of its revenue for the foreseeable future.
  • That creates a major risk:
    • Customers can switch to rival AI providers quickly.
    • Anthropic has not disclosed who the two customers are.
    • Investors may worry whether the company can turn current demand into stable, long-term contracts.

The Compute Problem: Anthropic’s Biggest Cost

  • Compute and infrastructure are Anthropic’s dominant expenses.
  • In the first half of 2026, it spent around $9 billion on compute, more than it spent in all of 2025.
  • Anthropic currently relies heavily on cloud providers like Google and Amazon, but says it has already tried to access more compute than those partners were willing to provide.
  • To keep growing, the company is:
    • Turning to smaller cloud providers
    • Building out its own infrastructure
    • Committing to large future capacity deals

Long-Term Spending Commitments

  • Anthropic has committed to roughly $450 billion in compute-related obligations over the next 7 to 10 years.
  • Including deals still being negotiated, total future compute spending could reach about $750 billion.
  • Most of those commitments are not cancelable, meaning Anthropic is locked in even if demand changes.

Major Risks Highlighted in the S-1

AI Safety and Liability

  • Roughly one-third of the S-1 reportedly deals with risk factors, including risks to humanity.
  • Anthropic warns that its models could:
    • Exhibit self-preserving behavior
    • Produce blackmail-like responses in testing
    • Enable sophisticated cyberattacks
  • The company says a catastrophic AI incident could create damages so large that insurance might not cover them.
  • Anthropic says it takes safety seriously and may delay model releases to reduce harm, but it also acknowledges uncertainty about preventing dangerous outcomes.

Data Center Backlash

  • Anthropic also flagged rising opposition to data centers, especially in the U.S.
  • If local or regulatory resistance slows infrastructure buildout, it could limit access to the compute Anthropic needs.
  • The company is exploring capacity outside the U.S. to diversify risk.
  • Building data centers is also getting more expensive because companies must make community-facing incentives and voluntary payments to ease concerns.

What Investors Will Likely Focus On

  • The biggest question for investors is whether Anthropic’s revenue can become durable and predictable.
  • Key watchpoints:
    • Can Anthropic secure multi-year customer commitments?
    • Will customers keep paying at today’s pace?
    • Can the company lock in enough compute without overcommitting?
    • Is AI demand strong enough to justify the company’s extraordinary spending?

Bottom Line

Anthropic’s draft S-1 paints a picture of a company with stratospheric growth, but also outsized dependence on a few customers and an enormous appetite for compute. The episode frames Anthropic as both a symbol of the AI boom and a test case for whether the economics of frontier AI can support the sky-high valuations now being discussed.