Why does Israel want to shift from being a high-tech leader to becoming an AI superpower?

Israel is seeking to reshape its economic image, moving away from the model widely known in the media as the “Startup Nation,” a reference to the high concentration of new technology companies, toward an economy more heavily reliant on artificial intelligence. Technological competition is no longer measured solely by the number of applications and startups, but by access to advanced algorithms, supercomputers, data centers, cutting-edge chips and vast datasets, as well as the talent capable of turning these resources into commercial products and security and military tools.

This direction is reflected in Israel’s National Artificial Intelligence Program, which places research and development, talent attraction, and the transformation of research and inventions into investable products and companies at the heart of a project aimed at strengthening Israel’s position among the world’s leading technological powers.

This project is not starting from scratch. The high-tech sector is a key pillar of the Israeli economy. According to the Knesset Research and Information Center, the sector accounted for around 10.4% of salaried workers in 2021, contributed about 15.3% of GDP and 54% of exports, while roughly a quarter of income tax revenues came from workers in the sector.

The shift toward artificial intelligence therefore represents an attempt to safeguard growth, exports and public revenues. The sector is developing within a network that connects universities, technology companies, research centers and development centers operated by global corporations, alongside growing state involvement in computing, data, education and infrastructure. This is transforming artificial intelligence from a commercial activity into an issue of national competitiveness.

The article argues that Israel’s attempt to move from a model built on a high concentration of startups to one more dependent on computing cannot rely on an abundance of engineers and ideas alone. It also requires continuous access to chips, data centers, advanced models and global capital. Under this model, the state is assuming a more direct role: rather than merely funding research, it is working to organize infrastructure, expand skills, create government demand and reduce the risks of dependence on external providers.

Accordingly, the Israel Innovation Authority’s 2026 strategy identifies computing capacity as one of the essential conditions for maintaining the competitiveness of AI companies, shifting the debate from “How many startups do we have?” to “Do we have the resources that allow these companies to stay in the race?”

From “Startup Nation” to an AI Economy

Israel’s economic model, built on high technology and innovation, depends heavily on global capital flows. However, political instability, the global shift toward artificial intelligence and the emergence of competitors to Israel in this field are increasingly limiting its ability to continue relying on this model alone.

Investment in the high-tech sector peaked at $24.9 billion in 2021, before falling to around $15 billion in 2022 and declining again in 2023. In 2025, companies raised around $14.6 billion, but the recovery was concentrated in large funding rounds, while financing for early-stage startups continued to decline. Israel Innovation Authority’s 2026 report indicates that the return of funding does not mean that the company base is expanding. Rather, it reflects the growing concentration of the market in a limited number of companies, primarily those working in artificial intelligence and its applications, that have reached advanced stages of growth and are better positioned to attract foreign investors.

Israeli company AI21 Labs provides a clear example. Founded to develop language models and enterprise-focused products, the company raised around $155 million in its Series C funding round in 2023, with participation from Google and Nvidia, bringing its valuation to $1.4 billion. The company’s figures illustrate that global capital is not merely investing in finished products; it is also betting on research teams capable of developing foundation models and competing with companies that possess far greater resources and computing power.

The same bet can be seen in infrastructure and acquisition deals. Nvidia built the Israel-1 supercomputer in Israel as a platform for testing and running generative AI applications, using advanced chips and networking technologies, significant parts of which were developed following its acquisition of Mellanox. In 2025, Google announced an agreement to acquire cloud security company Wiz for $32 billion, the largest acquisition in its history. These investments show that Israel’s technological value is no longer measured solely by the number of startups it produces, but also by its ability to provide the computing, security and data networks required by the global AI economy.

A second challenge is the growing relocation of activity and jobs abroad. In March 2026, only 62% of employees at privately held Israeli companies were working inside Israel, compared with around 69% in 2019, as activity expanded in the United States and the proportion of senior executives residing locally declined. The number of R&D workers also fell by around 3,500 in 2025, the first such decline in more than a decade. Israel Innovation Authority believes this may reflect changes in the nature of jobs amid the rise of AI tools, but it may also be linked to the relocation of talent and decision-making centers abroad amid war and political uncertainty.

Intel’s presence illustrates the depth of the Israeli economy’s ties to multinational corporations. Since entering the Israeli market in the 1970s, the company has expanded its manufacturing and R&D activities, employed more than 9,000 people, and recorded cumulative exports of around $86 billion through 2023, with its operations focused on processors, communications and AI solutions. Yet this integration involves a paradox: multinational companies provide financing, computing infrastructure and access to global markets, but they may also move intellectual property, management and decision-making centers abroad. Israel may therefore remain a center for knowledge production without always retaining its full economic value.

The third challenge is social. If artificial intelligence reduces demand for entry-level jobs, software testing and routine programming work, entering the sector may become more difficult for Arabs, Haredi Jews and residents of areas far from economic and technological centers, particularly Tel Aviv. These groups often lack access to the military and academic networks available to the technological elite. A 2025 Knesset study points to a relative decline in high-tech jobs as a share of the labor market, alongside the growing impact of AI on the nature and composition of jobs. Productivity may therefore rise while educational and professional pathways to high-paying jobs narrow, turning the technological divide into a broader class divide.

For several years, calls have therefore been growing in Israel to turn the country into an “AI superpower.” This ambition rests on a dense economic and technological base. Between 2012 and February 2023, the number of high-tech workers rose from around 274,000 to nearly 392,000. Average monthly wages in the sector reached 30,241 shekels, compared with 12,061 shekels across the economy as a whole, a difference of nearly 151%. Spending on research and development reached around 5.6% of GDP in 2021, compared with an OECD average of 2.7%. These indicators show that technological strength has not been a peripheral activity, but a pillar of high-wage job creation, exports and tax revenues.

The sector maintained its importance despite the war and slowing economic activity. High-tech output reached around 317 billion shekels in 2024, equivalent to 17.3% of GDP. In 2025, its share of exports reached a record 58%, and the sector contributed around half of economic growth, while employment approached 400,000 despite slower hiring compared with the average of the previous decade. According to the Israel Innovation Authority’s 2026 report, the sector’s recovery is no longer based on software alone, but is also linked to the rise of chips, servers, networking equipment and the computing infrastructure required for the AI era.

This is why the shift toward artificial intelligence carries direct economic significance. If cybersecurity—the protection of digital systems from intrusion—gave Israel an advantage in marketing security tools, generative systems capable of producing text, images and software, together with data-analysis technologies, extend this advantage into healthcare, finance, education, manufacturing and defense. Israel Innovation Authority says around 2,158 companies are engaged in AI-related research and development, including 199 branches of multinational corporations, alongside growing demand for specialists in data science, data engineering and machine-learning model operations.

Some key indicators of Israel’s technological infrastructure can be summarized as follows:

From Fragmented Initiatives to a State Project

Israel’s push toward artificial intelligence did not begin in 2026, but it gained greater urgency after years of incomplete plans. In 2019, recommendations were made for annual government investment of between one and two billion shekels to place Israel among the world’s five leading countries and replicate the experience of building its cybersecurity sector.

Artificial intelligence offers Israel a way to compensate for its limited population and territory by analyzing vast amounts of information, accelerating research and development, increasing the productivity of specialized labor and exporting digital products that do not require a large industrial base. According to the National Program for Artificial Intelligence, Israel ranks highly worldwide across several indicators and ranks second per capita in some of them. In addition, 492 Israeli AI companies received funding between 2013 and 2024, totaling nearly $15 billion, alongside a significant accumulation of research talent and participation in the development of advanced models.

Yet market dynamics moved faster than the state’s ability to build a coherent regulatory and institutional framework. In a 2024 report, the State Comptroller of Israel found that the country lacked a long-term strategy and a unified implementation plan, while the use of budgets allocated to high-performance computing and human-capital development had been slow, and Israel had fallen in several international rankings. What began as a problem of administrative coordination therefore evolved into a threat to the economic and technological standing, paving the way for the recommendations of the Nagel Committee and the creation of a more centralized national framework.

This gap reveals another paradox. Israel has companies and researchers with a global presence, yet it has been less capable of coordinating civilian policies, making data accessible and building public computing infrastructure that is relatively independent of foreign companies. This is reflected in the 2026 Digital Government Index, where Israel scored below the OECD average in data availability, accessibility and support for reuse, with scores of 0.42, 0.57 and 0.28 respectively. Having advanced companies, therefore, does not necessarily mean having interconnected digital institutions and databases that can be used jointly. Geopolitical power in the age of AI requires an integrated infrastructure of data, computing and education, rather than advanced but disconnected technological hubs within the military and private sector.

The institutional shift came in June 2026, when the Israeli government adopted a new national artificial intelligence program aimed at addressing gaps that had accumulated over several years. The National AI Directorate in the Prime Minister’s Office was tasked with coordinating its implementation. The plan includes establishing a national institute and laboratories to apply the technology in healthcare, education and transportation; expanding training programs; opening around ten university programs; offering incentives to bring experts back from abroad; and preparing a program to train nearly four million workers.

Implementation is estimated to cost around five billion shekels annually, reflecting AI’s transformation from a sector-specific innovation issue into a multi-ministry state program built around the parallel development of education, financing and government applications.

However, the rapid pace of implementation has sparked debate over priorities and the scale of spending. Researchers have warned against being drawn toward large-scale projects whose symbolic value may exceed their practical returns, while estimates suggest that some computing targets could push costs into the tens of billions of dollars over several years.

The business sector, by contrast, argues that delay has become more costly and is calling for human capital to be linked with commercial applications, infrastructure and education, while launching major national projects rather than relying on limited initiatives. This tension between fiscal caution and the industry’s push for rapid action reflects an awareness that the window for competition is narrow: success requires speed, but speed may lead to fragmented spending unless priorities are carefully aligned with needs and sectors where Israel has a comparative advantage.

Expanding Computing Sovereignty

Israel’s AI race is shifting from supporting software and startups toward building what is known as “computing sovereignty.” This refers to having sufficient domestic capacity to run advanced models without complete dependence on foreign data centers and services. Within this framework, the Knesset plan aims to secure around 100,000 processors dedicated to artificial intelligence, establish an advanced semiconductor plant and a quantum computer, classify server farms as national infrastructure, and accelerate procedures for their construction.

Implementing these plans, however, faces major challenges, particularly pressure on electricity, power grids and land. Requests to connect new server farms to the grid have reached around 20,000 megawatts, exceeding Israel’s current total electricity demand. The challenge is not limited to building servers; it also includes the high cost of computing and difficulties in obtaining advanced processors. A medium-sized Israeli company may spend around $200,000 annually, while waiting periods for processing units can reach six months because major global companies absorb a large share of supplies.

To address these constraints, Israel is examining international models for shared computing and regional agreements that would allow it to use capacity located outside its territory, amid intensifying competition from Europe, the Gulf and Asia. Computing sovereignty also extends to ensuring continuous access to advanced models, amid concerns that governments or companies may impose regulatory or geopolitical restrictions on their use. Recommendations therefore focus on building a basic domestic capacity, running models inside Israel when necessary, developing specialized models for Hebrew and public services, and diversifying suppliers. The aim is not complete self-sufficiency, but rather to reduce the risks of external dependence and prevent Israel’s strengths in software and talent from becoming a vulnerability.

The concept of sovereignty also extends to ensuring access to advanced models, as restrictions on their use for regulatory or geopolitical reasons have become a potential source of pressure. Recommendations therefore emphasize building a basic domestic capacity, running models inside Israel when necessary, developing specialized models for Hebrew and public services, diversifying suppliers, and incorporating guarantees of access to technology into strategic relationships. The objective is not full self-sufficiency, but rather managing external dependence in a way that prevents strengths in software and talent from becoming a source of vulnerability.

Conclusion

Israel’s attempt to shift from high-tech toward artificial intelligence is not merely a technological evolution, but an urgent necessity for maintaining economic growth and sustaining the “Israeli economic miracle,” as described in the Zionist narrative, which is organically tied to maintaining the factors that make Israel attractive domestically, regionally and internationally.

At the same time, current government efforts reflect a reorganization of the relationship between the state, universities, and local and global companies. Companies develop applications and models, while the state provides infrastructure, skills and public-sector demand, and multinational corporations provide chips, cloud services, markets and capital. As AI investment grows, the future of the Israeli economy is increasingly tied to its ability to retain ideas and talent within a system that has access to computing capacity, energy, financing and global markets.

Yet this rise remains marked by contradictions. The system is strong in research and development but weaker in coordination and data accessibility. It attracts major global companies but depends on them for chips, computing capacity and advanced models. At the same time, the expansion of server farms is placing growing pressure on electricity supplies and infrastructure.