人工智能企业的地理分布（英）

The geography of AI firms by Kumar Rishabh and Vatsala Shreeti Monetary and Economic Department April 2026 JEL classification: O33, C81, L86, F23, L16 Keywords: artificial intelligence, AI supply chain, firmgeography, AI measurement BISWorking Papers are written by members of the Monetary and EconomicDepartment of the Bank for International Settlements, and from time to time by othereconomists, and are published by the Bank. The papers are on subjects of topicalinterest and are technical in character. The views expressed in them are those of theirauthors and not necessarily the views of the BIS. This publication is available on the BIS website (www.bis.org). The geography of AI firms∗ Kumar Rishabh†Vatsala Shreeti‡ April 15, 2026 Abstract In this paper, we trace the geography and economic characteristics of firmsthat produce artificial intelligence (AI) products and services.Many economiesaround the world are evaluating their strategic priorities in AI, yet relatively littleis documented about the global distribution of AI production. We construct a newdatabase that identifies 1,246 AI-producing firms across 32 economies.We mapthese firms in each economy into the five layers of the AI supply chain: compute,cloud and related infrastructure, data tools, AI models and AI applications. Thebiggest markets for AI production are China and the US. Most economies specialiseonly in a few supply chain layers and many focus largely on compute. AI firms inall economies exhibit strong home bias in investment activity, with a focus ondownstream applications. Finally, we find that venture capital inflows are stronglycorrelated with the presence and density of AI firms in a given economy. Keywords:artificial intelligence, AI supply chain, firm geography, AI measure-mentJEL codes:O33, C81, L86, F23, L16 1Introduction Artificial intelligence (AI) is advancing rapidly and is being adopted across a wide rangeof sectors, from finance and healthcare to manufacturing and public administration. Agrowing body of research examines the implications of AI adoption for labour markets,productivity and macroeconomic outcomes.1 However, not much is known about thefirms thatproduceAI. Policymakers need to understand not only how AI is being used,but also where the capacity to develop and supply AI technologies resides and what theeconomic characteristics of AI-producing firms are. Many jurisdictions are increasinglyevaluating their comparative advantage within the AI supply chain and the trade-offsassociated with regulating AI firms and some are pursuing sovereign AI strategies. Yeta firm-level analysis of global AI production and its economic characteristics that caninform these discussions remains largely absent from the literature. As AI is becoming astrategic priority for governments around the world, this is an important research gap. In light of this gap, in this paper, we build and provide a new dataset of 1,246 AIproducer firms across 32 economies.We construct this dataset using a large languagemodel (LLM)-assisted classification and extensive manual verification of data sourcedfrom PitchBook. We use these data to identify the location of AI-producing firms andtrace their economic weight in their respective economies over time.We then analysethe economic and institutional factors that are correlated with a greater density of AI-producing firms. We also look at the variation in each economy’s AI supply chain anddocument the sub-markets in which different economies specialise: i) compute, ii) cloudand related infrastructure, iii) data tools, iv) AI models and v) AI applications. Finally,we analyse the investment linkages between AI firms across different economies.Theresulting map of AI firms and their economic and financial characteristics is not onlyuseful for understanding AI production within each economy, but can also inform researchseeking to quantify the cross-economy macroeconomic impact of AI. Consistent with previous work, we find that the United States (US) and China are the biggest AI markets in terms of both the number and valuation of AI firms (Frost,Shreeti, and Rishabh (2026)). Beyond this, the focus of our analysis is on documentingnew stylised facts. We provide four sets of results in line with our research questions. First, we find that AI firm density (ie the presence and number of AI firms) is posi-tively correlated with the size of the economy, share of research and development (R&D)expenditure in the economy and the size of venture capital (VC) inflows. VC inflows, inparticular, are positively correlated both with the probability of a jurisdiction having anAI firm and the number of AI firms in the jurisdiction. Second, the economic footprint of AI firms is especially large in the US, Korea andChinese Taipei. Compared with other economies, AI firms in the US and Korea accountfor significantly higher shares of market capitalisation at 40% in the US and 39% inKorea in 2025. The same pattern hold