Anthropic Projects Massive Revenue & Losses Amidst IPO Talk

Despite this impressive top-line growth, operating losses are expected to rise to $8.06 billion in 2025, up from $2.98 billion in the previous year. The company also reported a $42 billion net loss, which includes a substantial $34 billion accounting charge, and compute and infrastructure costs amounted to $7.33 billion in 2025, accounting for over half of all operating expenses.

Anthropic projects a massive $518 billion in future compute and cloud infrastructure obligations over the next few years.

Compute expenses alone have tripled compared to the preceding year, indicating aggressive investment in AI development.

The host identified three key areas of interest: the company's financials, the extensive list of risk factors, and the unanswered question of who leaked the prospectus and why. Analysts are scrutinizing the leaked prospectus, particularly the absence of 2026 financial figures, despite the fiscal year being three-quarters complete.

Speculation includes whether the leak was selective and its potential connection to concurrent industry events, such as OpenAI's recent developer day.

Some observers argue that parts of this loss, which could potentially be converted to Anthropic stock, represent actual monetary risks for the company.

This charge reflects an increase in the estimated value of financing that could eventually convert into Anthropic shares. Analysts questioned whether this charge is recurring or a one-time event that will not be repeated in future filings.

Nearly 25% of Anthropic's revenue is generated by just two customers, raising concerns about customer concentration.

The IPO prospectus, as reported by Reuters, highlights that many of the company's largest customers are not on long-term commitments, which means they could withdraw their spending at any moment. This is exacerbated by the fact that AI compute deals often have short, six-month windows, allowing for potential withdrawals.

Julie Hyman pointed out that risk factors occupy 80 pages of the 261-page filing, significantly more than the 48 pages dedicated to describing the actual business.

The document explicitly warns that advanced AI could pose 'catastrophic or existential risks to humanity.'

The hosts highlighted the unusual nature of an investment opportunity that simultaneously markets potential financial gain alongside the risk of ending the human race.

Anthropic's enterprise clients consistently demand next-generation AI models, creating pressure for continuous development.

The company's revenue is directly tied to a continuous and overlapping release cycle of new models, which is described as inherent to remaining at the frontier of AI development.

This rapid pace is deemed necessary despite the company's own warnings about the existential risks posed by advanced AI.

Tristan Slock's bottom-up analysis projects tech operating cash flow to reach $2.4 trillion by 2028, an increase of over $1.2 trillion.

Despite this significant growth in the tech sector, non-tech sectors in the S&P 500 are not showing commensurate revenue growth.

This raises concerns about how the massive tech build-out, particularly in AI infrastructure, will be paid for without corresponding revenue increases in other industries.

Ongoing debates persist regarding the sustainability of capital expenditure and financing within the AI sector, with figures like Ed Zitron raising concerns about capital and financing.

Despite these concerns, the investment community remains heavily committed to AI, pouring significant amounts of debt into funding AI infrastructure. This raises the critical question of who will be left holding the bag if AI investments do not yield the anticipated returns.