IMF: 23.1% of Israeli Workers Hold Jobs at High Risk of AI Displacement

23.1%. That is the share of Israel's workforce employed in occupations with high AI exposure but low complementarity — the combination the IMF flags as genuine displacement risk, not a productivity story. [Fact] If you work in sales, business administration, or even ICT, that category includes people who do what you do. The interesting part of this report is who lands on which side of the line — and why the answer cuts against most people's intuitions about which jobs AI threatens first.

The finding comes from "The Impact of Artificial Intelligence on Israel's Labor Market," an IMF Selected Issues Paper (2026/061) authored by Ece Ozge Emeksiz and published on July 13, 2026. [Fact] The paper applies the IMF's exposure-and-complementarity framework — the methodology developed in its 2023 cross-country work on AI and labor markets — to Israel's occupational structure. [Fact]

The headline is double-edged. Most Israeli workers are likely to come out ahead: they hold jobs where AI exposure is high but complementarity is also high, meaning AI tends to augment their work rather than replace it. [Fact] But roughly one in five workers sits in the high-exposure, low-complementarity zone. And at the other end of the spectrum, 32.8% of workers are in occupations with low AI exposure altogether — mostly manual work and interpersonal services that current AI systems barely touch. [Fact]

Who sits in the danger zone

The IMF identifies business and administration professionals, sales workers, ICT specialists, and some clerical occupations as the groups facing the greatest displacement pressure. [Fact] These jobs share a common structure: they are built around routine cognitive tasks — drafting documents, processing transactions, reconciling records, running standardized client interactions. That is precisely the task profile generative AI now performs cheaply and at scale.

By sector, the paper expects the largest workforce adjustments in ICT, finance, insurance, and professional services. [Fact] Notice what is absent from that list: no truck drivers, no factory floors, no construction sites. The displacement risk in Israel's economy concentrates in white-collar, high-value-added industries — the same industries that drove its growth over the past two decades.

The demographic pattern follows from the occupational one. Workers with tertiary education are heavily concentrated in AI-exposed occupations, and the IMF finds that younger workers are experiencing the shift fastest. [Fact] The stereotype of automation hitting older, less-educated workers first simply does not describe what AI is doing to a knowledge economy like Israel's.

This maps closely onto our own occupation-level data. The risk profiles we track for sales representatives, administrative assistants, and financial analysts show the same pattern the IMF describes: high task overlap with current AI capabilities, with the decisive variable being how much of the role involves judgment and relationships rather than routine processing.

The tech paradox: Israel's strength is also its exposure

ICT specialists appearing on a displacement-risk list is the counterintuitive finding here. Israel's technology sector is the engine of its economy, and its workers are among the most AI-literate anywhere.

But there is a counterargument the paper itself supports, and it deserves to be taken seriously: high exposure does not automatically mean job loss. The IMF finds Israel's labor market is less susceptible to AI-driven displacement than comparable European economies, precisely because complementarity is higher across its workforce. [Fact] ICT workers are also the group best positioned to convert AI into a productivity multiplier rather than a replacement. The split within tech occupations matters more than the sector average: routine implementation and testing work faces real substitution pressure, while system design, architecture, and problem framing remain firmly on the human side. Our data on software developers and computer systems analysts reflects exactly this internal divide.

So the honest reading is not "Israeli tech jobs are doomed." It is that Israel's tech sector will be reshaped earlier and faster than almost any other part of the economy — with large gains for workers who adapt and real losses for those whose roles were mostly routine.

Who is relatively safe — and why that is complicated

The 32.8% of Israeli workers in low-exposure occupations are concentrated in physical and interpersonal work: trades, caregiving, in-person services. [Fact] For public services such as education and healthcare, the IMF expects AI to improve efficiency rather than replace employees. [Claim]

Safe, then? In employment terms, largely yes — for now. But low exposure cuts both ways. These workers are also least positioned to capture AI-driven productivity gains, which raises a distributional question the paper's framework points toward: if AI boosts wages mainly in high-complementarity knowledge work, the gap between AI-augmented and AI-untouched occupations could widen even as displacement stays contained. [Estimate]

One limitation worth stating plainly: exposure and complementarity scores measure potential, not observed job losses. They are occupation-level estimates built on task descriptions, and they cannot capture how individual firms will actually deploy AI or how quickly workers will shift tasks within their jobs. The 23.1% figure describes risk, not a forecast of unemployment.

What the IMF wants Israel to do about it

The paper's central policy conclusion is a comprehensive lifelong learning strategy — reskilling for workers whose occupations shrink, upskilling for those whose jobs are being redefined, and mid-career training as a standing feature of the labor market rather than an emergency response. [Claim]

That prescription sounds generic until you connect it to the risk map above. The workers most at risk in Israel are educated professionals in mid-career, not entry-level or low-skill workers. Training systems in most countries are built for the young and the unemployed. A system that retrains a 45-year-old business administration professional into an adjacent, higher-complementarity role is a different institutional challenge — and that is the one the IMF is pointing at.

What this means if you are in one of these jobs

Three practical takeaways from the paper's logic.

First, audit your own task mix. The risk is not attached to your job title but to the share of your week spent on routine cognitive work. Shifting your role toward judgment, client relationships, and cross-functional coordination is the single most direct hedge.

Second, if you are in sales, administration, or finance operations, treat AI tools as mandatory literacy now. The IMF's framework says your outcome depends on whether AI complements your work — and complementarity is partly a choice you make by learning to direct these systems rather than compete with them.

Third, if you are mid-career, do not wait for national training programs to reach you. The paper's own conclusion is that Israel's institutions need to build mid-career retraining capacity that does not yet exist at scale. The workers who move early will move on their own initiative.

The report's bottom line is genuinely more hopeful than the 23.1% headline suggests: most Israeli workers stand to gain from AI adoption. But "most" is doing quiet work in that sentence. For one worker in five, the difference between augmentation and displacement will be decided by what they — and Israel's policymakers — do in the next few years.

Sources

• IMF Selected Issues Paper 2026/061 — The Impact of Artificial Intelligence on Israel's Labor Market
• Full paper PDF (imf.org)

This article was produced with AI-assisted analysis of the IMF Selected Issues Paper 2026/061. All statistics are attributed to the IMF report; occupation-level risk data linked above comes from our own ONET-based analysis.*