Quick answer: AI will reshape the global economy by automating cognitive tasks across nearly every industry, lifting productivity (Goldman Sachs estimates a potential 7% — about $7 trillion — rise in global GDP over 10 years), shifting income from labour toward capital, and widening gaps between AI-ready and AI-exposed workers, firms and countries. The IMF estimates that roughly 40% of jobs worldwide — and 60% in advanced economies — are exposed to AI. The direction of change is clear; the speed and distribution of the gains are what remain uncertain.

The numbers at a glance: $15.7T added to the global economy by 2030 (PwC) · about 7% GDP boost over 10 years (Goldman Sachs) · $2.6–4.4T in annual value from generative AI (McKinsey) · 300M jobs exposed to automation (Goldman Sachs) · 40% of jobs worldwide and 60% in advanced economies exposed (IMF).

What "economic impact of AI" means

The economic impact of AI is how it changes productivity, employment, wages, prices and the distribution of wealth across an economy. Economists treat AI as a general-purpose technology — like electricity or the internet — so its effects ripple through nearly every sector rather than staying in tech. The difference this time: AI is the first technology to automate cognitive work at scale — writing, coding, analysis, design, support — which is why its footprint could be larger and arrive faster than past automation waves.

1. Jobs: task automation, not just job loss

AI automates tasks within jobs more than it eliminates whole jobs. Most roles are bundles of dozens of tasks; AI absorbs some while humans keep the rest, and jobs get reorganised around what people do best.

— Goldman Sachs estimates 300 million full-time jobs are exposed to automation globally, with roughly two-thirds of US occupations exposed to some degree.

— Of exposed occupations, only about a quarter to half of the workload is automatable — the rest still needs a human.

— Historically, displacement has been offset by new work: 60% of today's jobs didn't exist in 1940.

What's different now is that knowledge work is hit first, the transition may be faster, and the adjustment period — retraining, relocating, re-skilling — is where the real pain lands.

2. Productivity and GDP: a potential growth surge

If AI broadly augments cognitive work, it could lift output significantly:

— Goldman Sachs: +7% global GDP (about $7 trillion) and +1.5 percentage points of productivity growth, over 10 years.

— PwC: up to $15.7 trillion added to the global economy by 2030.

— McKinsey: $2.6–4.4 trillion in annual value from generative AI.

The catch is the productivity paradox. As with computers in the 1980s, gains can lag investment by years until firms redesign their processes around the technology. Diffusion, not invention, is the bottleneck.

3. Wages and inequality: the central risk

The most likely danger from AI is not mass unemployment — it is widening inequality.

— Capital may capture more value than labour, because AI is itself a form of capital that substitutes for cognitive work.

— Wage polarisation: some workers become far more productive and valuable; others find their skills commoditised.

— A possible upside: several studies show AI can lift lower-skilled workers toward expert-level output on specific tasks, which could compress some skill gaps. The net effect is still being measured.

The IMF warns that AI is likely to worsen inequality overall unless offset by deliberate policy.

4. Market and power concentration

The economics of frontier AI — enormous compute, data and energy costs — favour a small number of large firms. That raises antitrust and dependency concerns as critical infrastructure concentrates, and geopolitical stakes as nations leading in compute, energy, talent and deployment gain durable advantage. The counterforce is open-source and smaller, efficient models that push value back toward broader access.

5. Sector-by-sector transformation

— Services (the bulk of advanced economies): exposed to automation at scale for the first time.

— Healthcare: faster diagnosis, drug discovery and administrative savings.

— Education: personalised tutoring, and pressure on traditional models.

— Finance and law: document analysis, research and decision support automated.

— Energy: data centres become a major new source of electricity demand, reshaping power markets and capital spending.

6. Prices, deflation and measurement

When content, code and analysis become near-free to produce, costs fall in affected sectors — a deflationary pressure. But much AI value arrives as free or cheap services that don't show up cleanly in GDP, which makes the real impact hard to track and to tax.

7. Policy, taxation and the safety net

If displacement outpaces job creation, expect intensifying debate over retraining and re-skilling at national scale, wage insurance and stronger safety nets, new tax models (from "robot/AI taxes" to rethinking how automation-created value is captured), and recurring — and contested — proposals for universal basic income.

Which jobs are most and least exposed?

Most exposed (cognitive, routine-digital): data entry, basic customer support, paralegal research, bookkeeping, routine content production, junior analysis and coding tasks.

Least exposed (physical, relational or high-judgment): skilled trades, hands-on healthcare, complex negotiation, creative direction, and roles that need physical presence or deep human trust.

The IMF's exposure by economy: about 60% of jobs in advanced economies (roughly half negatively affected, half augmented), 40% in emerging markets, and 26% in low-income countries. The gap is double-edged: advanced economies face more disruption but are better placed to capture the gains.

FAQ

Will AI cause mass unemployment?

Most economists think widespread inequality is a bigger risk than mass unemployment. AI tends to automate tasks rather than whole jobs, and historically new technologies have created new categories of work even as they displaced old ones. The danger is the speed and unevenness of the transition.

How much will AI add to the global economy?

Estimates vary by scope and timeframe. PwC projects up to $15.7 trillion by 2030, Goldman Sachs estimates a roughly 7% (about $7 trillion) GDP boost over 10 years, and McKinsey estimates $2.6–4.4 trillion in annual value from generative AI specifically.

How many jobs will AI affect?

Goldman Sachs estimates 300 million full-time jobs are exposed to automation globally, and the IMF estimates about 40% of jobs worldwide are exposed — but exposure means tasks could change, not that jobs will necessarily disappear.

Will AI make inequality worse?

Likely yes, without policy intervention. The IMF warns AI may worsen inequality because it shifts value toward capital and high-skill workers — though some studies show it can also lift lower-skilled workers on specific tasks.

The bottom line

The direction of AI's economic impact is clear: more automation of cognitive work, higher potential productivity, and growing distributional stress. The magnitude and speed are genuinely uncertain — and they hinge less on the technology than on choices about education, competition, energy and taxation. The economies and individuals that thrive won't necessarily have the best models; they'll be the ones that adapt their skills, processes and policies fastest.

Sources

— Goldman Sachs: Generative AI could raise global GDP by 7%

— IMF (2024): Gen-AI: Artificial Intelligence and the Future of Work

— McKinsey: The economic potential of generative AI

— PwC: AI to add $15.7 trillion to the global economy by 2030