Anthropic goes public amid an AI slump: the problem is no longer the model, it's the cost of building it

By The Washington Post (AI & Tech Brief, by Dan Gallagher) · November 2, 2026.

Anthropic is pressing ahead with its IPO (initial public offering) despite a sharp drop in AI-linked stocks. According to the Post's analysis, the company has fast-growing revenue and a huge cash position that give it a strong case for facing a difficult market.

The reason for the jitters, according to the same piece, is the cost of building AI. Results from Google, Microsoft and Amazon were overshadowed by massive capital expenditures (capex) and by signs that AI may be disrupting other parts of their businesses.

The figure the article highlights most: Anthropic has nearly $900 billion in commitments over the next five to eight years. The Post adds that the industry's growing reliance on debt makes the AI buildout a potential risk to the credit market. The available material is a brief excerpt, so it includes no breakdown of revenue, valuation or IPO timeline; we do not treat them as known.

Our take: the debate has shifted. Two years ago the question was whether the models were useful; today it is whether the business can bear its own bill. These are different questions, and the second is more financial than technological. That the problem is cost rather than demand is, paradoxically, a sign that the technology works well enough to demand investment on this scale.

The real short-term risk is not that AI is useless, but that it is financed with leverage and that a correction drags down lenders and weaker companies. In general, infrastructure bubbles (railroads, fiber optics) left painful bankruptcies, but also assets that society later put to use. That analogy does not guarantee the outcome, but it helps avoid mistaking a stock market correction for a failure of the technology.

The winners in this context are those with the cash and revenue momentum to hold out; the losers, players that depend on cheap debt and on the market continuing to believe. For jobs and society, the nuance matters: financial consolidation could concentrate even more power in a few labs, something we already monitor in our ongoing coverage.

In the long run, our thesis has not changed: if today's investment leads to more effective medicine and an abundance of resources, the price paid, even with turbulence, will have been worth it. But the transition will be uneven, and for exposed investors and workers, considerably more uncomfortable than the optimistic narrative usually admits.