One could devote an entire blog nowadays to AI-related financial reporting matters…

Well, no doubt plenty of people are already doing that. A big source of content for them the foreseeable future will be this, as reported here in the New York Times:

• Anthropic, a five-year-old company whose artificial intelligence tools have made it a world leader in the technology, on Monday fully laid out the business behind its big ambitions for the first time.
• In a financial prospectus for what could be the largest ever initial public offering, the San Francisco company revealed that sales of its A.I. products like the Claude chatbot to other businesses had grown at an astonishing clip. But it also said that it was losing money and planned to spend at an even faster rate as it invests heavily in A.I. infrastructure to support that growth.
• Anthropic’s revenue was $16.3 billion in the first half of the year and its net loss was $80.7 billion, driven largely by a one-time charge. Stunningly, the company has committed up to $807 billion in contracts for data centers, chips and computing power, many of which cannot be canceled, over the next 10 years, according to a tally by The New York Times.

Anthropic’s own expression of the economic opportunity it represents evades any concept of normal quantification:

• If  Powerful AI is directed at solving humanity’s hardest problems and the risks are managed responsibly, the potential is extraordinary. The possibilities include compressing decades of scientific progress into years; accelerating drug discovery, disease detection, and the delivery of personalized therapy; providing individually tailored education and expert financial and legal advice; advancing power grids and battery design for greater efficiency and reliability; and extending to the developing world capabilities that took wealthy nations generations to build
• Frontier AI is transforming cognitive work—the application of knowledge, reasoning, and judgment to produce economic value. We believe it will make humans far more capable and expand the world’s overall capacity for that work, including accelerating scientific discovery and helping businesses grow…While it is still early to forecast frontier AI’s full impact across the economy, International Data Corporation (“IDC”) estimates that investment in AI solutions and services will yield a cumulative global economic impact of $22.5 trillion by 2031. We believe the impact could be even greater.

At the same time of course, the risk factors are voluminous, for example: “Market adoption may be slower than anticipated due to factors such as uncertainty regarding the cost, performance, reliability, and perceived value of AI solutions; increased regulation; concerns about security, privacy, legal and regulatory compliance, or ethical considerations; technological challenges; level of effort and expense associated with integration efforts; competition from alternative or emerging technologies; the availability or adoption of complementary infrastructure and workflows; or reductions in corporate spending due to macroeconomic conditions. Moreover, demand for frontier models may decline if customers determine that less advanced, lower-cost models are sufficient for their particular use cases.” And going back to that New York Times article:

• Anthropic’s filing was unusual in that it warned its products could bring about disasters, including loss of control of the technology and authoritarian abuses. “To build A.I. is to harness cognition itself, and anything so powerful can be used for ill as well as good,” Dr. Amodei and the other founders wrote.
• One jaw-dropping risk factor: Advanced A.I. systems could “pose catastrophic or existential risks to humanity,” the company said.

Against that dizzying backdrop, commentators and markets are becoming increasingly aware of how much is financially on the line here. Alphabet, Meta, and Microsoft Are Hiding $3 Trillion in Debt on the AI Boom’s Hidden Ledger, announced the headline of a Yahoo Finance article:

• In July, Nikkei estimated that Microsoft, Amazon, Alphabet, Meta Platforms and Oracle were carrying roughly $1.65 trillion in obligations that don’t appear as debt on their balance sheets — mostly future spending tied to data centers, chips, and power. That was startling enough. But a Wall Street Journal analysis published this month found the real figure across nine companies — those five plus Nvidia, Broadcom, Advanced Micro Devices, and SpaceX, — sits closer to $3 trillion, based on filings mostly current through June 2026.
• That’s not a rounding error. It’s the difference between a concerning trend and a systemic one. The Journal’s breakdown shows roughly $1.9 trillion in purchase commitments — contracts to buy chips, servers, and hardware from suppliers — plus another $1.2 trillion in leases that have been signed but haven’t yet commenced, meaning they still sit outside reported balance-sheet liabilities…

In all key available respects, the disclosure might appear to be just about as transparent as it could reasonably be: only the most willful investor could fail to grasp the magnitude of the operational and financial uncertainties (as I write, the stock of most of the companies mentioned has taken a recent dip). As such, heavily AI-dependent stocks should likely be shunned for now by all but the most flamboyantly risk-embracing investors. But of course, their reach, and potentially calamitous downsides, will be far greater than that…

The opinions expressed are solely those of the author.