This post was originally published on this site.
AI-specific resources like chips, data, and infrastructure can become a source of global inequality, the International Monetary Fund (IMF) mentioned in a paper it published in April this year. The organisation highlighted that policy actions like the US’s plan to restrict exports of advanced AI chips on a company and country basis, and the EU’s instruction to its member states to review outbound investments in AI and semiconductors, suggest the growing importance of digital infrastructure.
While the advanced economies (like the US) have access to these digital resources, emerging markets and low-income countries do not, constraining their capacity to innovate and compete in sectors exposed to AI.
“Uncertainties about future trade and technology restrictions add to these challenges, potentially limiting opportunities for technology transfers and collaboration. As a result, access to AI-specific resources—chips, data, and infrastructure—risks becoming a further source of global inequality, cementing the advantage of economies that can produce or readily acquire the critical inputs needed to harness AI’s transformative power,” the IMF explains.
Why it matters:
The IMF’s working paper highlights the growing geopolitical and economic impact of AI access. This is relevant in the current context, where the US is considering tariffs on semiconductors, according to a Reuters report. According to a TechBrew report, if the price of semiconductors were to increase as a result of these tariffs, the cost would get passed on to AI companies like Microsoft and Google, and eventually to the end user. This, in turn, could impact AI adoption and innovation.
Critical elements for AI capacity:
The paper suggests that a country’s AI capacity depends on three elements—access to technologies, data, and infrastructure (as explained above), the degree of AI exposure, and AI preparedness.
The degree of AI exposure:
This depends on how extensively AI affects a country’s workforce and industries. Jobs serve as the primary conduit through which the effects of AI manifest in industries and economies. The IMF says that, given the different occupational structures across countries, 60% of jobs in advanced economies have high exposure, while 42% of jobs in emerging economies and 26% of jobs in low-income countries have high AI exposure. AI exposure in jobs can be both positive and negative. For advanced economies, it could mean a larger and more immediate potential for productivity gains, along with the risk of labor displacement.
Speaking about AI-related job losses in June last year, IMF Deputy Managing Director Geeta Gopinath mentioned that during previous waves of automation, companies held on to employees when they were making profits. However, she added that “the extent to which automation could replace humans only becomes fully visible during or immediately after a downturn.”
AI preparedness:
This refers to the readiness of a country’s institutions, digital infrastructure, workforce, and industries to adopt AI. The IMF notes that historical episodes of technological change have shown that a country’s ability to benefit from new technologies depends heavily on its readiness. As per the IMF’s AI Preparedness Index (released in June last year), India ranks 72nd among 174 countries in terms of AI readiness.
The IMF suggests that in countries where foundational preparedness for AI is weak (emerging markets and low-income countries), investment in digital infrastructure and human capital should be prioritized to reap early gains from AI while paving the way for second-generation preparedness. It further mentions that governments should prioritize investing in areas such as fundamental research, necessary infrastructure, and applications where social returns are high, such as healthcare, education, and government administration.
The international body also notes that the recent advancement of large language models through efficient algorithms by DeepSeek “provides a glimmer of optimism to AI underperformers.” It explains that DeepSeek’s breakthrough in creating an AI model at a relatively lower expense disrupts the assumption that cutting-edge AI systems need a high budget. “This breakthrough in the production of a more efficient LLM shows that technology leapfrogging cannot be dismissed, even under limited access to more advanced hardware components,” the IMF says. It adds that open-source code and relevant data could help less-developed emerging economies and low-income countries apply AI to address their own problems.
Also read:
Support our journalism: