Macro Markets and Machines_The Economic and Market Transformation Driven by AI_GWM report - Flipbook - Page 33
Macro, Markets, and Machines
November 2025
3.
Access
Access is the final major determinant for cross-country divergences in would-be benefits from AI. The IMF’s
framework considers that countries may have unequal access to equipment or IP, limiting their ability to
innovate and remain competitive in sectors that would capitalize on AI technology. Geopolitical or national
security issues or export restrictions would be among the factors that constrain access, particularly for
emerging markets or lower-income countries that would also not have the capability to build up domestically.
China’s build-out of DeepSeek at a much lower cost than G7-born AI models suggests this divergence
between developing and emerging markets could be less pronounced, albeit mostly for larger nations with
greater resources.
These barriers exacerbate the inequality in economic outcomes under the IMF’s baseline scenarios. The IMF’s
high-TFP simulations that impose AI access restrictions on emerging market economies (excluding China) show
that while the United States, advanced economies, and China see only a marginal drag on GDP gains versus
baseline, emerging market nations experience about 40% smaller output gains versus baseline. Latin American
economies would see slightly less than a 2% boost to GDP over 10 years, compared with the United States with a
5.4% gain (Exhibit 25).
GDP increase versus baseline
10 years ahead
Exhibit 25 – AI Access Can Limit Already-Low GDP Gains from AI Adoption
6%
5%
4%
3%
2%
1%
0%
ROW
LIC
EMA
Limited AI access
EML
CHI
EUS
OAD
USA
World
Baseline AI access
Notes: CHI = China. EMA = Emerging Market Economies Asia, Central Asia, Russia, etc. EML = Emerging Market Economies Latin America, Middle
East, Africa, etc. EUS = EU and Switzerland. LIC = Low-Income Countries. OAD = Other Advanced Economies.
Sources: IMF; Scotia Wealth Management.
Section 4.5: Global GDP Outcomes Build Off Productivity Differences
With the estimates on the impact of AI adoption on productivity as outlined in the previous section, the IMF
finds that global GDP would be around 2.5 p.p. higher over baseline in a high-TFP scenario over the first
five years, rising to 4 p.p. at the 10-year mark. In a low-TFP scenario, gains are estimated at 0.8 p.p. and 1.3 p.p.
five and 10 years out, respectively (Exhibit 26).
Global benefits from AI are unevenly distributed, however, as the United States and European Union see GDP
boosts of 5.4 p.p. and 4.4 p.p., respectively, after 10 years in a high-TFP scenario, compared with 3.2 p.p. in
emerging markets excluding Asia and only 2.7 p.p. in low-income countries. The IMF points to shortcomings in AI
preparedness and exposure, highlighting the relatively higher share of employment in agricultural and manual
labour that fails to make use of AI technology.
Scotia Wealth Management
32
