Anthropic’s IPO prospectus, filed with the SEC on 26 October 2026, says: “We believe the work that knowledge workers perform represented more than $40 trillion in economic value globally in 2025.” The S-1 adds that, with frontier AI’s potential to make those workers more productive, Anthropic’s near-term opportunity “is to expand this value and accelerate economic growth.”
The Wall Street Journal ran an explainer on 28 October headlined “See How Anthropic Arrived at Its Mind-Boggling $40 Trillion Market Opportunity.” The filing itself says how the number is built, and it is a wage estimate.
Short explainer · AnthropicAnthropic and the GSA#8217;s S-1 numbers
A vertical news Short narrated from the published text of this piece and related FourWeekMBA pieces.
Business Pill · AI agents explained
An evergreen Business Engineer explainer on AI agents. It explains the general idea and does not discuss Anthropic’s S-1.
The key insight: As we read it, Anthropic’s $40 trillion is a wage bill, not a sales forecast: the S-1 builds it from US Bureau of Labor Statistics wages scaled to the world, and calls expanding that value its near-term opportunity.
More than $40 trillion (Anthropic’s estimate of knowledge work’s 2025 economic value), more than $118 trillion (2025 global GDP, World Bank, as cited in the S-1) and about 34% (our arithmetic).
The S-1 describes the figure as its own estimate: “Certain estimates in this prospectus regarding global knowledge workers’ wages, which we refer to as the annual economic value represented by global knowledge workers, are our own, derived from third-party data.”
The filing lays out the method in four steps. Anthropic takes 2025 US wage data by occupation from the Bureau of Labor Statistics and adds up the salaries of the workers it says it believes represent knowledge workers. It estimates 2025 global wages from International Labor Organization data on labour income as a share of global GDP, applied to 2025 global GDP. It then takes the ratio of total US wages to that global wages figure and applies it to the US knowledge-worker wage total.
The S-1 does not list which occupations it counts as knowledge work, and it does not give the US or global wage totals behind the estimate.
Dollar figures in Anthropic’s S-1 (filed 26 October 2026), in trillions: global GDP of more than $118 trillion in 2025 (World Bank, as cited), knowledge workers’ economic value of more than $40 trillion in 2025 (Anthropic’s own wage-based estimate), information technology spend of $4.3 trillion and cloud spend of $1.9 trillion in 2025, and pharmaceutical spend expected to exceed $2 trillion in 2026 (IQVIA, as cited). Different measures, shown for scale and not summed; the one-third share is our arithmetic.
The Other Numbers in the Section
Anthropic’s S-1 (filed 26 October 2026) puts the economic value of global knowledge workers’ work at more than $40 trillion in 2025, its own wage-based estimate, and cites global GDP of more than $118 trillion (World Bank). The one-third share is our arithmetic.
The same section cites International Data Corporation (IDC), which, the S-1 says, “estimates that investment in AI solutions and services will yield a cumulative global economic impact of $22.5 trillion by 2031.” Anthropic adds: “We believe the impact could be even greater.”
The S-1 says information technology spend grew from approximately $1.2 trillion in 2006 to $4.3 trillion in 2025, a CAGR of approximately 7%, as cloud computing spend grew from zero to $1.9 trillion. It says global GDP grew from approximately $34 trillion in 2000 to more than $118 trillion in 2025, according to the World Bank.
The filing also says annual global pharmaceutical spend is expected to exceed $2 trillion in 2026, citing IQVIA, and that Claude is used in drug discovery and development by customers, biotechnology partners and Anthropic’s own researchers.
By our arithmetic, $40 trillion is about 34% of the $118 trillion global GDP figure the S-1 cites. The two numbers measure different things: one is wages for a subset of workers, the other is total output.
How Others Have Framed It
PYMNTS wrote on 26 October that “Anthropic estimates its total addressable market at more than $40 trillion, based on the economic value generated by knowledge workers globally in 2025.” It added that the figure represents the “economic activity Anthropic hopes to serve, rather than a forecast of AI spending.”
The Wall Street Journal headline calls it a market opportunity. The S-1 wording is narrower: it calls the $40 trillion the economic value of knowledge workers’ work, and calls expanding that value its near-term opportunity.
As we read it, the method matters more than the size: the S-1 does not name the occupations it counts, so the figure cannot be rebuilt from the filing alone.
As we read it, the filing frames Claude against what knowledge work is worth, not against what software spend is: the S-1 sets $40 trillion beside $4.3 trillion of information technology spend in 2025.
As we read it, the outside framing runs ahead of the filing: PYMNTS called the figure a total addressable market, while the S-1 calls it economic value and treats expanding it as the opportunity.
Anthropic, Form S-1
“We believe the work that knowledge workers perform represented more than $40 trillion in economic value globally in 2025”
Three Implications
FOR INVESTORS READING THE S-1 The S-1 says its market data involves assumptions and limitations and cautions readers not to give undue weight to such estimates.
FOR SOFTWARE VENDORS The S-1 sets knowledge workers’ economic value of more than $40 trillion beside $4.3 trillion of 2025 information technology spend, both as stated in the filing.
FOR EMPLOYERS The S-1 says Anthropic’s near-term opportunity is to expand the value of knowledge work as frontier AI makes those workers more productive.
The framework puts it this way: “AGaaS charges for the execution of outcomes.”
As we read it, sizing the opportunity by wages rather than by software budgets is the outcome-pricing view applied to a prospectus. The framework was written in March 2026 and does not discuss Anthropic’s S-1.
What Is Not Established
The S-1 does not say what share of the $40 trillion Anthropic expects to earn, over what period, or how it relates to its own revenue. It does not name the occupations it counts as knowledge workers.
The prospectus also warns that its market data “involves a number of assumptions and limitations, and you are cautioned not to give undue weight to such estimates.” We could not read the Wall Street Journal article in full; only its headline is cited here.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Anthropic’s S-1 puts the economic value of global knowledge work at more than $40 trillion in 2025, an estimate it builds from US wage data scaled to the world. The filing calls expanding that value its near-term opportunity; outlets have described the same figure as Anthropic’s market opportunity.
A note on sourcing. Figures are from Anthropic’s Form S-1 filed with the SEC on 26 October 2026 (market-opportunity passages and the Industry, Market, and Other Data section, read in full), and from PYMNTS (26 October 2026). The $40 trillion figure is Anthropic’s own estimate and was not independently verified. The Wall Street Journal article was not read beyond its headline. This is not investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
