8JUN

IMF: AI jobs run 3.6% lower after 5 years

2 min read

11:04UTC

An IMF Staff Discussion Note estimated that employment in AI-vulnerable occupations runs 3.6% lower after five years in high-AI-demand regions, with entry-level roles most exposed.

← Back to Jobs report says fine, layoff report says no Jump to analysis ↓

EconomicDeveloping

Key takeaway

AI-exposed jobs run 3.6% lower after five years, hitting entry-level workers hardest.

The International Monetary Fund (IMF), the Washington-based lender of last resort to its member states, estimated in a research note published in January 2026 (SDN 2026/001) that employment in AI-vulnerable occupations runs 3.6% lower after five years in high-AI-demand regions than in low-demand ones ¹. The note found entry-level roles carry the highest AI exposure of any career stage.

A 3.6% gap that opens over five years stays small in any single month's payroll print, so it does not register as a shock; it accumulates as a slow erosion that only a regional comparison reveals. The IMF isolated it by contrasting places where AI demand is high against those where it is low, controlling for the broader economy.

The study supplies the structural backbone for the youth-unemployment figures landing this week . It identifies the mechanism, suppressed junior hiring in the most AI-exposed regions, that the UK and US headline figures show as outcomes without naming the cause. Because the effect builds over years and concentrates on workers entering the labour market, the full cost will appear long after the policy decisions that might have addressed it.

Deep Analysis

In plain English

The International Monetary Fund, the global financial organisation that lends to countries in economic difficulty and publishes economic research, published a study in January 2026 comparing regions where AI is used heavily with regions where it is used less. They found that in the high-AI regions, jobs in the most AI-exposed occupations were 3.6% lower after five years than in comparable low-AI regions. For context: the IMF counted entry-level roles as the most exposed. So the study suggests that over five years, regions where companies adopted AI heavily ended up with about 36 fewer jobs in AI-exposed occupations for every 1,000 that existed. Researchers disagree about whether this five-year decline will continue to grow, level off, or reverse as new roles emerge around AI. But the finding is one of the few large-scale attempts to measure the actual employment effect, as opposed to predicting it.

What could happen next?

Consequence
If the IMF's -3.6% five-year regional estimate is causal, entry-level hiring in AI-exposed occupations will decline a further 1-2 percentage points in high-adoption regions over 2026-2028.

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: The 2015 Acemoglu-Restrepo research on industrial robots used the same regional high-vs-low-exposure methodology and found -0.2% employment per robot per 1,000 workers, a figure subsequently contested on endogeneity grounds by subsequent studies.

The IMF's -3.6% applies the same regional-comparison design. Acemoglu and Restrepo's estimate survived most robustness checks even after challenge; whether the IMF's AI estimate similarly survives will become clear as the five-year window completes.

Counter-parallel: Oxford Economics' January 2026 study found no consistent acceleration in productivity that would confirm labour substitution at the IMF's implied scale, a contradiction that could mean either the IMF's regional comparison is capturing something other than AI displacement, or productivity gains are accumulating in ways aggregate statistics miss.

Consensus view: The IMF's SDN 2026/001 authors, Era Dabla-Norris and team, and the OECD's Going Digital research group both describe the -3.6% five-year employment figure as mechanistically conservative: it captures only direct occupation-level substitution and does not count suppressed hiring of new graduates into AI-exposed roles.

The Hosseini Maasoum and Lichtinger SSRN paper found entry-level postings fell 15% in AI-adopting firms while senior roles stayed flat, meaning the IMF's -3.6% understates total cohort displacement by excluding the graduate-entry suppression channel.

Counter-view: The New York Federal Reserve's May 2026 occupation-level job-posting research found minimal evidence AI has directly reduced labour demand since ChatGPT's launch, and identified the relative decline in AI-exposed postings as predating December 2022.

This challenges whether the IMF's regional comparison correctly attributes the -3.6% to AI rather than to pre-existing sectoral trends in high-AI-demand regions (typically high-cost metropolitan areas with above-average housing-driven labour supply constraints).

Key tension: Whether the five-year -3.6% employment gap reflects AI substitution (the IMF's framing) or pre-existing structural economic geography (the NY Fed's challenge), a distinction that determines whether the figure represents a causal prediction or a correlation.

Sources:International Monetary Fund

Mentions:ChatGPT →Federal Reserve →OECD →SSRN →Washington →Oxford Economics →

First Reported In

Update #12 · Jobs report says fine, layoff report says no

International Monetary Fund· 8 Jun 2026

Read original →

Different Perspectives

Office for National Statistics

Deferred its Transformed Labour Force Survey beyond November 2027 and disclosed a May 2026 telephone-collection failure. The ONS carries no AI-attribution layer at all, so Britain sits outside this month's cohort of measuring states by its own admission.

Uber India, Swiggy, Zomato and Urban Company

Named as respondents after the Karnataka High Court extended the interim welfare-fee deposit arrangement under the state's gig-worker welfare law to Uber India on 28 July, joining the other platforms already under the same order. The companies are contesting the underlying law while complying with the interim deposit terms.

Kenya State Department for ICT and the Digital Economy

Its draft AI policy, open for consultation to 4 August, proposes a pay floor for data-annotation work, where Kenyan annotators earn $1.46 to $3.74 an hour against $21 to $27 in the US, on figures relayed by the trade outlet WeeTracker. Kenya is legislating on AI labour even though the World Bank rates it among the least exposed economies.

ARAN and Italian public-sector unions

Signed the CCNL Funzioni Centrali 2025-2027 on 6 August, the first Italian national contract with a dedicated AI Title, barring fully automated employment decisions without meaningful human intervention and requiring advance union notice of AI deployment. The unions secured this through bargaining rather than waiting for legislation.

US employers reporting to Challenger, Gray & Christmas

Named artificial intelligence as the leading stated cause of job cuts for a fifth consecutive month in July, at 33% of that month's total, even as the overall cut count fell 27%. Employers kept citing AI as the reason even as scrutiny of the attribution rose.

Bank for International Settlements

Bulletin 130 reports a 0.75 percentage point average unemployment rise across high-AIPI countries between 2023 and 2025, while its own footnote 2 states the index is strongly correlated with employment shares in AI-exposed sectors it is used to predict. The bulletin calls the productivity payoff uncertain and uneven.