Goldman projects $685 billion in revenue for Anthropic in 2030: the figure matters less than who will govern that power

By Yahoo Finance · November 10, 2026.

Goldman Sachs projects that Anthropic could generate $685 billion in revenue in 2030, Yahoo Finance reports. That figure is more than 130 times the company's 2025 sales. The news comes as the company, according to the same source, prepares for an IPO at a $1.4 trillion valuation that could happen as soon as next week.

The projection comes alongside equally enormous spending: according to the same report, Anthropic plans to invest about $900 billion in building out its AI infrastructure. In other words, the bank is not only talking about revenue: it describes a company that would burn capital on a scale comparable to the world's largest industrial companies.

The article adds a political angle. According to Yahoo Finance, a company of that size would make Dario Amodei, Anthropic's CEO, one of the most influential executives in the world, and would strengthen institutions aligned with his views on effective altruism and existential risk. The Trump administration has condemned effective altruism, and Mark Zuckerberg (Meta) and Jensen Huang (Nvidia) have called Amodei a "doomer," again according to the source. Note that the available material is a short piece: it does not detail Goldman's methodology or assumptions.

**Our take.** A four-year projection for a company growing at this pace is, above all, a scenario, not a fact. Growing 130-fold in five years requires everything to go right: sustained demand, margins that can withstand $900 billion in spending and a favorable regulatory environment. It is best read as a sign of what the market is willing to believe, not as a forecast that will come true as stated.

What is really interesting is not the number but the consequence the piece itself points to: when a frontier AI lab becomes one of the world's largest companies, its governance stops being an internal matter. As we have argued before about IPO prospectuses, the S-1 becomes a lab's most detailed public record, and the securities regulator ends up exercising, in practice, a form of AI oversight. In addition, with the IPO, the risk of this technology can reach household savings.

There is also an irony worth highlighting. The company criticized as "doomer" is the one preparing for a $1.4 trillion IPO. This suggests the market does not penalize caution on safety but treats it as part of the product: trust is a commercial asset in the enterprise segment. In our view, that is good news, because it aligns economic incentives with prudence. The condition is that this prudence is demonstrated through external evaluations and audits, not just rhetoric.

In the short term, the uncomfortable part: a concentration of capital and power of this size in a few labs worsens inequality between those who control the technology and those who bear its effects as workers, and it fuels the politicization of the debate, as the clash with the White House shows. In the long term, however, that same financial muscle is what can fund the science that could bring us closer to eradicating diseases and to shared abundance. Whether that happens depends less on Goldman's figure than on the rules used to govern whoever reaches it.