The global economy may be holding steady, but the International Monetary Fund’s managing director is sounding an alarm about who gets left behind in the process. Speaking in Bangkok, where Thailand is hosting the IMF–World Bank Annual Meetings from October 12 to 18, Kristalina Georgieva described artificial intelligence as a “tsunami” for jobs — one that could sustain growth at the top while quietly dismantling the entry points that young workers have always relied upon.
Georgieva told a panel on Monday that AI investment is doing real work to keep the world economy moving, growing at roughly 3% even as energy supply shocks drag on global output. Countries building AI infrastructure — many of them across Asia, including Thailand — are contributing meaningfully to that momentum. But a growing economy, she was careful to note, does not automatically translate into better career prospects for the people entering it.
The IMF estimates that AI will affect approximately 40% of jobs worldwide, a figure that climbs to 60% in advanced economies. Some of those jobs will be enhanced by automation; others will be downgraded or eliminated outright. The distribution of that disruption, however, is far from random, and Georgieva was direct about who faces the sharpest exposure.
“The people to worry about first and foremost are young people who are coming into the labour market,” she said. Historically, new workers have entered through routine and repetitive roles — unglamorous, perhaps, but functional as a foundation. Those positions offered a foothold, a place to learn and accumulate the experience needed to move upward. AI is now eroding exactly that kind of work, and with it, the conventional path to professional advancement.
The metaphor Georgieva used — a career ladder with its first rung pulled away — echoes a concern that has been building across labour market research. LinkedIn executive Aneesh Raman has described the new career landscape as a “climbing wall,” one that demands lateral moves and continuous adaptation rather than the linear promotions workers once expected. Human resources strategists have similarly argued that hiring practices must shift toward soft skills and adaptive capacity, rather than fixed technical competencies that AI can replicate or replace.
What makes Georgieva’s intervention notable is the institutional weight she brings to these questions. The IMF is not typically in the business of lamenting the social texture of work; it measures output, debt, and fiscal balance. Yet here was its managing director pressing governments to rethink the very architecture of taxation and education in response to a structural shift in how labour and capital relate to one another.
“Traditionally, governments would generate revenues by taxing mostly people,” she said. “Well, if AI changes the structure of the labour market and there is less labour, more capital — what do we tax?” The question is not rhetorical. It points toward a genuine fiscal challenge that democratic governments will need to confront as automation reshapes the tax base that funds public services, including the social programmes that cushion economic transitions.
On education, Georgieva was equally pointed. The old model — acquire a skill, hold it for a lifetime — no longer fits the pace of technological change. Lifelong learning, she argued, is now a “paramount necessity,” not an aspiration reserved for the professionally ambitious. Governments, she said, carry a serious responsibility to redesign how education functions in an AI-shaped world, and to do so before the gap between what schools produce and what economies require becomes impossible to close.
The Bangkok panel offered no tidy solutions, and Georgieva did not pretend otherwise. What she offered instead was a clear-eyed account of the stakes: a technology that can simultaneously lift aggregate growth and hollow out the foundations of individual economic mobility. For young Canadians and young workers everywhere, the challenge is not simply to adapt to AI — it is to ensure that the institutions built to support them adapt first.
