IMF AI report 2026: deployment and disruption
The IMF AI report 2026, part of the Fund’s annual report, has a separate section on how AI changes the economy, the labor market and financial stability. By the Fund’s estimate, investment in AI technology added about 0.5 percentage points to US GDP growth in 2025. Below are the main numbers of the report and of the IMF source papers, Russia’s score in the AI Preparedness Index and the lessons for a company.
Key figures
0.5 pp
added to US GDP growth in 2025 by investment in AI-related technology
2 trillion dollars
is how much private AI investment worldwide may exceed in 2026, by some outside estimates the IMF cites
41,2%
of US companies in the information industry used AI in the six months before the February 2026 survey; 31.7% in professional services and 12.7% in manufacturing, where growth to 17% is expected
Figure 1.1, p. 16 (PDF 18); values on the section’s web page
1 in 10
of job ads in advanced economies require at least one new skill, about half as often in emerging markets; job ads with AI skills offer higher pay
3,6%
lower employment after five years in US regions with growing demand for AI skills, for occupations with high AI exposure and low complementarity, which is about 30% of all employment
0,56
is Russia’s score in the IMF AI Preparedness Index on a 0–1 scale (data for 2023); Singapore has 0.80, the US 0.77 and China 0.64
The report in brief
- Published by
- International Monetary Fund
- Published on
- 23 September 2026
- Original title
- The 2026 IMF Annual Report: Navigating a Precarious World. In Focus: AI: Deployment and Disruption
- Length and language
- 92 pages, the AI section on pp. 15–16, in English
- Checked
- 16 October 2026
- Next edition
- yearly, before the IMF and World Bank annual meetings; the report for financial year 2027 is expected in autumn 2027
“technology investments related to AI added an estimated 0.5 percentage point to US GDP growth”p. 15 (PDF 17), In Focus: AI: Deployment and Disruption
How the report was made
This is the IMF Executive Board’s annual report for the financial year that ended on 30 April 2026 (p. 1). The AI section summarizes the Fund’s research and outside estimates (pp. 15–16). Figure 1.1 is based on the US Census Bureau business survey for November 2025 and February 2026, with IMF calculations (p. 16).
The numbers on skills and employment come from the IMF source paper SDN/2026/001, January 2026 (p. 2, pp. 19–20). The paper estimates demand for new skills from online job ads and compares US regions with different growth in such demand. The AI Preparedness Index covers 174 countries on a scale from 0 to 1, with data for 2023 (IMF DataMapper, indicator AI_PI, method section).
What the report does not measure
The AI section takes two pages and has almost no numbers. The IMF takes the estimate of 2 trillion dollars of investment from outside sources and does not name them (p. 15). Data by industry exists only for the US. The report states its labor market conclusions in words: workers with AI skills earn more, but regions with many such job ads see no overall gain in jobs (p. 15).
The AI Preparedness Index has not been updated since 2023. The IMF itself asks readers to treat it as a guide, not a ranking, so the Fund does not publish country places (AIPI Note, p. 1).
What this means for a company in Russia
Russia in the report: The annual report has no AI data on Russia. In the IMF AI Preparedness Index (data for 2023), Russia has 0.56 on a 0–1 scale; the average for emerging markets is 0.46, and for advanced economies 0.68. The IMF does not publish country places. Russia is not in the sample of paper SDN/2026/001 on skills and job ads (Annex Table 1.2, p. 29).
- By outside estimates, AI investment may exceed 2 trillion dollars in 2026 (p. 15), and the IMF directly names the risk that costly investment, more and more often on debt, will not pay back (pp. 15–16). For a company, the rule is simple: first a verifiable result on one process, then the budget. That is how a pilot on one process works.
- In advanced economies, one job ad in ten already requires a new skill (SDN, p. 2), and per the IMF, the losers are mid-skill workers whose work is highly exposed to automation (p. 15). Start with the staff whose tasks are changing: they need corporate training before new tools.
- In the US, 41.2% of companies in the information industry and 12.7% in manufacturing use AI (Figure 1.1). The report has no such data for Russia, and the US numbers do not carry over directly because the methods differ. Our study of Russian industry readiness shows the state of things by industry in Russia.
What to compare with
- AI in US companies: the BTOS 2026 survey: the primary source of the US industry data, from the same Census Bureau survey
- ILO: generative AIgenerative AI. Generative AI is a family of neural networks that create new content: text, images, music, video and program code, from a description in plain language. and work 2025: the exposure of occupations to generative AI by the ILO method
- World Bank: AI foundations 2025: AI foundations by country group: data centersAI data center. AI data centers are buildings with thousands of servers on graphics cards and other accelerators, with strong power supply and cooling. Neural networks are trained and run there., models, skills
Sources and dates
- IMF, Annual Report 2026, In Focus: AI: Deployment and Disruption, checked 16 October 2026imf.org
- IMF, The 2026 IMF Annual Report: Navigating a Precarious World (PDF), checked 16 October 2026imf.org
- IMF, Bridging Skill Gaps for the Future: New Jobs Creation in the AI Age (SDN/2026/001), checked 16 October 2026imf.org
- IMF DataMapper, AI Preparedness Index, Russian Federation, checked 16 October 2026imf.org