The AI bomb ticking under Israel's economic miracle - and what could stop it from going off
According to an international study by Deutsche Bank, Israel’s economy is among the world’s most exposed to AI disruption, with a highly AI-exposed labor market and high-tech service exports accounting for 8% of GDP. How Israel navigates the AI revolution will determine whether it emerges as a major loser - or winner
What happens to Israel's economy if AI slashes the cost of the very things it sells to the world — code, financial analysis, R&D? That's no longer a hypothetical. A new Deutsche Bank report finds that Israel, more than any other developed economy, has bet its growth on a single basket: exporting knowledge and technology services. If the AI revolution erodes the value of that basket, the fallout won't stop at tech layoffs - it could mean fewer dollars flowing into the economy, pressure on the shekel, lower wages, weaker consumption, a cooling real estate market, and a hit to state tax revenue.
But the same exposure cuts both ways: if Israel plays its cards right, that dependence could just as easily make it one of the biggest winners of the AI era. Of the thirty economies the bank studied, none has a wider gap between the best-case and worst-case outcome than Israel - which is exactly why the question "how much of the economy can AI replace" shuld be the most urgent one facing Israeli policymakers right now.
Israel is one of the most exposed economies in the world to disruption of services by AI, according to the report recently published. The report examined two things: what share of the workforce is employed in AI-based professions, and to what extent the country's foreign trade relies on exports of AI-based services. According to the study, Israel and the United Kingdom are two economies in which an unusual combination exists: a labor market highly exposed to AI, combined with a high dependence of foreign trade on exports of these services.
In most European countries a high share of the workforce is exposed to AI-based professions, but without their foreign trade being overly dependent on them. In India and the Philippines, a small share of the workforce is employed in such professions, but the countries' external balance relies heavily on exports of these services. Very few developed economies are as dependent as Israel on the sectors that the AI revolution is about to directly affect. Out of the thirty economies examined, Israel ranks first in net exports (exports minus imports) of services the bank classifies as AI-exposed - about 8% of GDP (followed by the UK, with about 7% of GDP).
The reason is clear: over two decades, the Israeli economy underwent a dramatic structural shift. It transformed from an economy that exports more goods into one that exports more knowledge services, software, and R&D. According to the study, Israel is among the countries where service exports compensate for goods imports. It appears alongside the UK, India, the Philippines, and Romania as an economy that is a net exporter of services but a net importer of goods, with Israel's case being more extreme.
About 55% of all Israeli service exports in the past year were concentrated in the programming sector and related activities. Israel is one of the largest net exporters of the services Deutsche Bank defines as "AI-affected," such as computing and tech-based services in fields like business services, finance, intellectual property, and insurance.
What will happen to the economy's growth model if the price, quantity, and nature of production of knowledge services in the world changes? As of today, high-tech accounts for 18.3% of Israel's output, but is responsible for 50% of its growth.
In the negative scenario, Deutsche Bank's economists suggest imagining a world in which intensive use of AI dramatically lowers the marginal cost of the knowledge economy. Code, financial analysis, consulting, design, translation, or research that today take hundreds of work hours could in the future be done in a fraction of the time. In such a situation, a large share of the economic value could shift to whoever holds the chips, the data centers, and the models themselves, while countries that make their living from exporting knowledge could be hurt.
For Israel, then, the damage could be far greater than mere layoffs in the high-tech sector. It would mean less revenue from exporting technological knowledge, fewer dollars entering the economy, and pressure on the shekel to depreciate. At the same time, average wages would be lower, households would consume less, real estate purchases would decline, tax revenues would fall, and capital gains would shrink. All of this translates into damage to growth.
High-tech in Israel employs about 10% of the economy's workers and is responsible for more than a third of income tax revenues. A shock to the sector therefore affects the entire economy. And this is not a future scenario: AI is already beginning to show up in Israel's employment data. In 2025, the number of R&D workers in high-tech fell by about 3,500 - the first decline in a decade - with their share dropping from 51% to 49% of the sector's employees. In high-tech and software services, R&D positions fell by 3.5%, even though output across the sector as a whole rose sharply. The Innovation Authority notes that the change in role structure may also be related to efficiency gains achieved through AI.
Moreover, in a survey conducted this year by the Innovation Authority among 210 companies employing about 130,000 workers, the share of companies saying that AI adoption led them to cut back on hiring jumped from 3% to 10% within six months, and half of the companies planning layoffs said AI has an influence on that decision.
On the other hand, the positive scenario could be no less dramatic. Deutsche Bank sketches an alternative scenario in which an Israeli engineer, aided by AI, is able to produce three to five times more per work hour; a software company develops a new product within three months instead of a year; a startup reaches sales of hundreds of millions of dollars with hundreds of employees instead of thousands. In this scenario, Israel's unusual structure turns it into one of the biggest winners.
Countries that already hold a comparative advantage in services may be able to leverage their existing knowledge and expertise to increase export volumes faster than prices fall. In simple terms, if software becomes far cheaper to produce but the world consumes far more software, Israel could come out ahead. And this is already happening: in 2025, high-tech output grew in real terms by 8.2%, while the number of high-tech workers grew by only 2.5%, and output per worker rose by about 6% to roughly 827,000 shekels. In other words, Israeli high-tech is already producing much more without significantly increasing headcount.
Deutsche Bank's researchers stress that it is impossible to know in advance which scenario will materialize, but their contribution lies in mapping the range between the positive and negative scenarios. In Israel's case, that range is the widest of all. The important question for Israel, therefore, is not how much of its labor is exposed to AI, but how much of Israel's value comes from work that can be replaced, and how much comes from ownership of technology that can be amplified. This, too, is measurable — and the news on that front is less encouraging.
Using Anthropic's data, Deutsche Bank examined whether AI is being used mainly to augment workers or to replace them. According to the report, in Israel usage skews relatively more toward automation, which is consistent with the high concentration of service exports in programming. Whichever government comes next will need to track this variable closely.
On the other hand, Israel has another layer of exposure that Deutsche Bank barely factors into its model: capital. The bank focused on the labor market and the external balance, and even noted that future research should add the equity-markets channel. In Israel's case, this channel cannot be left out. Nasdaq, foreign investment in high-tech, exits, capital raises, and institutional portfolios connect the Israeli economy to the global tech revolution through the financial account as well. AI could therefore hurt Israeli exports while raising the value of tech assets worldwide - or, in another scenario, hurt both domestic revenues and the value of tech companies alike.
So the big story is not whether ChatGPT or Claude will replace a programmer. Over twenty years, Israel successfully built an economic model based on turning expensive human capital into technological services and products that the world is willing to pay a great deal of money for. Now comes a technology whose central promise is to radically cheapen that very same human capital - while simultaneously and radically increasing its productivity. According to the Deutsche Bank report, very few countries have put so many eggs in this basket, which is why Israel could be one of the biggest winners from the AI revolution - or, without the right navigation, could find that the technology that changed the world is also changing the model on which the Israeli economic miracle was built.
