When most companies prepare for an IPO, investors focus on the usual questions.
How fast is revenue growing?
Is the business profitable?
What valuation makes sense?
Anthropic’s upcoming public filing is different.
The company has already confirmed that it confidentially submitted draft IPO paperwork to the SEC. The filing itself remains private, but enough information has emerged through funding announcements, investor reports, and disclosures from other companies to paint a surprisingly detailed picture of what could become one of the largest and most closely watched technology listings in history.
And if the publicly reported numbers are even close to reality, Anthropic may be operating at a scale that would have seemed impossible for an AI startup just a few years ago.
The Revenue Growth Is Almost Difficult to Believe
The most eye-catching figure is the reported revenue trajectory.
According to investor materials that were reported publicly, Anthropic generated approximately $4.8 billion in revenue during the first quarter of 2026.
More remarkably, the company reportedly projected $10.9 billion in second-quarter revenue, along with roughly $559 million in operating profit.
If those projections prove accurate, Anthropic would achieve something many observers thought was years away for large AI models: meaningful operating profitability.
That matters because one of the biggest debates surrounding generative AI has never been demand.
Demand is obvious.
The debate has always been whether AI companies can make enough money to justify the extraordinary infrastructure costs required to train and serve these models.
For years, skeptics argued that AI would become a race where revenue grows rapidly but expenses grow even faster.
Anthropic’s reported numbers suggest there may be a path to profitability, although the company itself has reportedly warned investors that profitability may not be consistent because of future compute obligations.
That caveat is important.
A profitable quarter does not necessarily mean a sustainably profitable business.
The Real Story May Be the Commitments
Revenue is impressive.
The commitments are staggering.
Anthropic’s eventual S-1 will likely reveal one of the largest collections of infrastructure obligations ever disclosed by a technology company.
The xAI Agreement
One of the most surprising disclosures came not from Anthropic, but from xAI’s corporate ecosystem.
Public reports indicate Anthropic has committed approximately $1.25 billion per month for exclusive access to the Colossus supercomputing facility.
That translates to roughly:
- $15 billion annually
- More than $40 billion over the life of the agreement
- Access to over 220,000 NVIDIA GPUs
- Massive power and computing capacity dedicated to AI development
Think about that for a moment.
Most technology companies spend years building toward billion-dollar annual contracts.
Anthropic is reportedly committing more than a billion dollars every month just for computing resources.
The AWS Relationship
Then there is Amazon.
Anthropic’s relationship with Amazon has evolved far beyond a typical strategic investment.
Publicly reported agreements indicate:
- More than $100 billion in AWS spending commitments over a decade
- Access to Amazon’s Trainium AI chips
- Multi-gigawatt infrastructure capacity
- Billions of dollars in additional Amazon investment tied to commercial milestones
This creates a fascinating dynamic.
Amazon is simultaneously:
- Investor
- Infrastructure provider
- Strategic partner
- Beneficiary of Anthropic’s growth
The relationship is deeper than most traditional technology partnerships.
The TPU Financing Structure
Another major question revolves around the reported TPU leasing arrangement involving Google hardware and large institutional investors.
Reports suggest a financing structure worth roughly $36 billion that allows specialized AI chips to be acquired and leased back to Anthropic through a special-purpose vehicle.
Why does this matter?
Because accounting treatment can dramatically change how investors interpret a company’s financial health.
Depending on how these arrangements are classified, billions of dollars of obligations may appear very differently on financial statements.
This is exactly the type of issue that public-market investors will scrutinize once audited disclosures become available.
The AI Economics Debate Is About to Get Real
For the past three years, conversations about AI economics have largely relied on estimates.
Anthropic’s filing could change that.
Investors are desperate for answers to a few fundamental questions:
- What are actual AI gross margins?
- How much does model training cost?
- How expensive is inference at scale?
- How much revenue comes from enterprise customers?
- How concentrated is the customer base?
These questions matter because the future valuation of every major AI company depends on the answers.
Bullish investors believe AI platforms will eventually resemble software businesses with strong margins and recurring revenue.
Bearish investors believe infrastructure costs will permanently limit profitability.
Anthropic’s audited financials could provide the first truly comprehensive look at which side is closer to reality.
The Governance Structure May Be Even More Unusual Than The Financials
Most IPOs ask investors to trust founders.
Anthropic asks investors to trust a governance structure.
The company operates as a public benefit corporation and is overseen by a unique Long-Term Benefit Trust.
The trust consists of five financially disinterested members whose authority grows over time.
Eventually, the trust gains the power to elect a majority of the board.
That is highly unusual.
In traditional public companies:
- Shareholders elect directors.
- Voting power generally follows ownership.
Anthropic’s structure introduces a different philosophy.
The organization was designed around the idea that advanced AI systems require oversight mechanisms that extend beyond maximizing shareholder value.
Supporters argue this protects the company’s mission.
Critics may question whether public investors will be comfortable owning shares in a company where governance authority is intentionally separated from economic ownership.
Regardless of where investors stand, it represents one of the most significant governance experiments ever attempted at this scale.
Why This IPO Matters Beyond Anthropic
The importance of Anthropic’s eventual filing extends far beyond one company.
The document could become a benchmark for evaluating the entire AI industry.
Investors will compare Anthropic’s disclosures against:
- OpenAI
- xAI
- Amazon
- Microsoft
- Future AI startups preparing for public markets
For the first time, markets may get a clearer view into the economics behind frontier AI development.
The questions are no longer whether AI is useful.
That debate has largely been settled.
The questions now are:
- Can AI businesses generate durable profits?
- How much infrastructure spending is required to stay competitive?
- Do the largest players gain sustainable advantages from scale?
- Can investors earn attractive returns despite enormous capital requirements?
Anthropic’s filing may not answer every question.
But it could provide the most detailed glimpse yet into the financial engine powering the AI revolution.
And for investors trying to understand where artificial intelligence is headed next, that may be the most important disclosure of all.