The annual meetings of the International Monetary Fund and World Bank are being held in Bangkok this year against a backdrop of equal parts relief and dread.
In her curtain-raising speech, IMF head Kristalina Georgieva said the global economy had navigated through serial shocks with “remarkable agility,” and that the meeting would focus on energy prices, high government debt, and the rise of AI.
The relief comes from the resilience of the global economy in the face of a war in the Middle East that has continued far longer than was expected at the spring meetings held in Washington in April.
The dread comes from worries that the war in the Gulf (and other conflicts) are continuing; that the energy buffers protecting the global economy are being eroded; that AI, which is seemingly offsetting the risks to global growth right now, is a potential source of immense job losses if it works and financial shocks if it does not work; and that rising global bond yields are threatening not just the economy but also fiscal solvency in many countries.
On the positive side, in the most recent edition of the World Economic Outlook (WEO), the IMF noted that its projection for global growth had remained essentially unchanged at 3.1% for 2026 and 3.4% for 2027. The Fund cited the flexibility of oil markets that saw increased production outside the Persian Gulf, the use of alternate transportation routes out of the region, and increased recourse to alternative energy sources as some of the reasons for the resilience.
The other mitigating factor, an extraordinary increase in spending on AI, has also helped cushion the global economy but unevenly and with costs and uncertainties. The technology boom has seen a ramp-up in global investment, but this has been geographically concentrated, with U.S. firms expected to spend close to a trillion dollars this year alone on AI infrastructure. And the corollary to all this investment has been a huge jump in imports of key semiconductor inputs from East Asia, particularly Taiwan and Japan.
The flip side of this boom has been not just an increase in Asian trade surpluses and U.S. trade deficits (often a source of friction), but also a rise in the interest rates of long-dated government debt across many countries as the scale of the buildout forces a competition for savers’ wallets. The AI investment boom has come alongside renewed inflationary pressures that have been exacerbated by the war in the Gulf, and increasing concern about the debt trajectory of G7 members like the U.S. and France.
The rise in interest rates also stems from worries about the independence of the Federal Reserve in the case of the U.S., where President Trump is continuing to pressure the institution via social media and via efforts to remove key personnel.
In France, worries about rising deficits and debt are intersecting with concerns over the 2027 election, as two major contenders, Marine Le Pen on the right and Jean-Luc Mélenchon on the left seem unwilling to comply with various policy constraints imposed by the European Commission and the European Central Bank.
One outcome of these intersecting factors is that global growth in the WEO is projected to be much more uneven. Growth in emerging markets is expected to be 3.7% in 2026, down just 0.1 percentage points versus the estimates in April, but this includes an India that has proven more resilient seeing a 0.6 percentage point boost to 7.0%, while the Middle East and North Africa see a 1 percentage point drop, led by a dramatic 5.0 percentage point fall in Saudi growth estimates (vs. April) to a 3.3% contraction.
While WEO global projections have barely changed over the last six months, the distribution has. In April, 54 economies were expected to account for 95% of global growth, but now only 32 economies will. Along similar lines, growth prospects have been downgraded for 47% of the economies covered by the Fund. The main factors driving these outcomes are the relative extent (and direction) of exposure to wars in the Middle East and Europe on the one hand and to updraft from the AI investment boom on the other.
For oil-importing countries like South Korea and Taiwan that are also major producers of key equipment required for the AI boom, the impulse from AI-related exports has given their economies a boost that has overpowered the negative impact of higher oil prices. Singapore, Malaysia, and Vietnam have also seen similar effects, but at a smaller scale. In a similar twist, oil exporters in the Global South that are distant from the Persian Gulf have seen growth upgrades versus April expectations, but oil importers with less exposure to the tech boom could face rougher waters. Accordingly, of Africa’s two biggest economies, Nigeria is seen as benefiting, and South Africa as being held back.
These are not the only concerns. 2026-27 is projected to be a Super El Nino year, and one that comes when farmers around the world are already facing stresses from higher diesel and fertilizer prices. The phenomenon is associated with heavier rainfall in some areas (like Northern East Africa and South India) and with drier conditions in others (Southern Africa and North India), raising the specter of higher food prices that hit the poor the hardest.
A more widespread worry is the impact that higher interest rates on U.S. Treasury debt could have on financing conditions for countries of the Global South. However, it should be noted that increasing anxieties about the AI growth trajectory have brought down U.S. Treasury yields as of this writing).
The problem is that many borrowers in the Global South, and particularly those that run trade deficits, are forced to keep interest rates higher than they might otherwise be in order to attract capital when U.S. rates are high, even if their budget deficits and gross debt are much lower than those of developed countries. Current interest rates in Brazil are roughly 9% above the rate of inflation, for example. This, in turn, can slow growth and worsen the debt trajectory.
Finally, there is AI itself – the center of the current boom, and the node of multiple anxieties. One set of medium-term worries is about the extent to which AI will eliminate jobs in services, particularly at the entry-level, a risk highlighted at the IMF meetings by Singapore President Tharman Shanmugaratnam. This is a global issue, but especially prominent in countries with a large youth cohort seeking upward mobility through education, particularly in Africa and South Asia.
One such country is India, where the Gen-Z cockroach movement got its start protesting against leaks and other irregularities in the nationwide competitive exams that govern entry into education and employment.
And if the socio-economic consequences of AI that works “too well” could lead to one set of bad outcomes, the financial consequences if AI does not prove as profitable as hoped could lead to another set of problems. Gartner, a global technology advisory firm (known among other things for having coined the concept of the hype cycle) expects an additional $6 trillion of AI spending through 2030, but that $3.5 trillion of AI assets could be financially impaired (i.e., not generating sufficient returns) by that date.
The IMF notes that interlocking financing patterns that connect vendors of AI hardware, the builders of data centers that use their chips, and the AI lab enterprises could lead to unanticipated stresses. As if these were not enough, the same IMF publication also highlights the prospect of AI “working well” but at “the wrong things” as the technology enables ever more sophisticated cyberattacks on critical and financial infrastructure.
The world may have dodged a bullet thus far this year, but the shots are still flying.