IMF, OECD Warn Poorer Countries Risk Falling Behind in AI Economy

Poorer countries, smaller businesses and younger workers risk missing out on the economic benefits of artificial intelligence unless access to electricity, digital infrastructure and relevant skills improves, officials from the International Monetary Fund (IMF) and the Organisation for Economic Co-operation and Development (OECD) have warned.

The warning comes as heavy investment in AI chips and computing infrastructure helps support global economic growth, raising questions about whether companies can generate sufficient returns from their spending. While AI could increase productivity across the world, countries with weaker infrastructure and fewer skilled workers face greater barriers to adopting the technology.

Era Dabla-Norris, deputy director of the IMF’s Fiscal Affairs Department, told the Reuters Global Markets Forum that “the critical question” for developing economies is whether they can adopt and spread AI throughout their economies.

The IMF estimates that effective AI adoption could add up to 0.5% to annual global economic growth over the medium term. Dabla-Norris cautioned, however, that evidence of productivity gains has so far fallen short of the expectations surrounding the technology.

The OECD estimates that AI could add between 0.4% and 1.2% to annual productivity growth across its member countries. Its chief economist, Stefano Scarpetta, said workplace adoption remains behind consumer use, with limited integration into core business functions.

“There is a large gap between exposure and actual use,” Scarpetta said.

The concerns reflect a divide in countries’ ability to turn access to AI tools into measurable economic benefits. An OECD working paper published in December 2025 found that low-income and lower-middle-income countries could benefit less from AI because they have smaller knowledge-intensive service sectors, where many productivity gains are expected to occur.

The paper also identified inadequate digital infrastructure, limited access to financing, workforce education and skills gaps, and underdeveloped regulatory frameworks as barriers to adoption. Although younger populations and international knowledge transfers could help developing economies, the OECD warned that existing structural weaknesses could outweigh those advantages.

Africa faces several of these constraints. In a July 2026 study on AI in Sub-Saharan Africa, the IMF identified unreliable electricity, limited digital infrastructure, shortages of technical skills and gaps in regulatory and institutional capacity as obstacles to wider adoption.

The IMF study estimated that cumulative productivity gains in Sub-Saharan Africa over the next decade could range from 0.2% under current conditions to 2.1% under a scenario of high AI adoption. These figures are scenario estimates rather than observed gains, and the report links the potential benefits to improvements in the conditions needed for adoption.

Employment is another concern. The IMF estimates that around 40% of jobs worldwide could be exposed to AI, although exposure does not necessarily mean a job will be eliminated. The technology can also assist workers with existing tasks and change the skills employers require.

Dabla-Norris said younger and entry-level workers could be particularly vulnerable if AI takes over tasks through which employees traditionally gain experience and develop skills for more advanced roles.