Ensuring Israel’s AI leadership: Are we prepared for the next revolution? - opinion

Artificial intelligence is rapidly becoming a technology of the present, not merely an innovation of the future. It is quickly emerging as a central infrastructure on which the economy, science, industry, and national security of the coming decades will be built.

For Israel, this is not simply an academic question. It is a strategic, and perhaps even existential, issue affecting both economic strength and technological superiority.

Many countries recognize the importance of artificial intelligence and are currently investing enormous sums, far beyond what Israel can afford, to build a competitive advantage in the field.

The committee’s central recommendation, which was implemented at record speed, unusually fast for the government sector, was the establishment of a strong national body capable of leading the field and coordinating efforts among the government, academia, the defense establishment, and industry.

The idea was straightforward: artificial intelligence is not an issue that belongs to a single government ministry. It affects nearly every area, including defense, healthcare, education, transportation, energy, industry, and science.

Israel therefore needs a central coordinating body that can see the full picture, set priorities, and ensure investments complement rather than duplicate one another.

The Directorate’s strategic plan, published in 2026, follows the main directions outlined by the committee.

However, establishing the body itself is only the first step. The real test will be its ability to translate decisions into a multi-year program supported by an appropriate budget, as the committee recommended, with clear timelines, defined responsibilities, and measurable indicators of success.

Computers can be purchased. Data centers can be built and supplied with electricity. Billions can be invested, as several countries seeking leadership in this field are doing.

But an entire generation of highly qualified researchers, scientists, and engineers cannot be created overnight.

For this reason, one of the committee’s most important recommendations focuses on investment in human capital.

That investment must begin at an early age, through educating children and teenagers, continue by increasing the number of students and researchers at every level, and include creating conditions that encourage Israeli experts who moved abroad to return, particularly those who left because of insufficient academic infrastructure in Israel.

Israel must also expand the circle of people who use artificial intelligence at every level.

Software engineers should not be the only professionals who understand and use AI. Doctors, teachers, industrial professionals, economists, lawyers, government employees, and managers will all need to know how to use the technology and, of course, understand its risks.

This is not a one or two-year effort. It is about building the human infrastructure for an entire generation.

One of the central conclusions of the Nagel Committee is that becoming a significant player in artificial intelligence requires adequate computing power.

The committee recommended establishing national computing infrastructure and a supercomputer with at least 60,000 next-generation GPUs, about half of which would go to academia and support relevant startups.

At first glance, this may seem like a purely technological project. In reality, it is a fundamental question of national independence.

A country without independent access to significant computing power becomes dependent on those who possess it. When academic or defense research, sensitive technologies, or strategically important systems are involved, that dependence can have significant consequences.

National computing infrastructure is therefore not merely another infrastructure project. It is part of Israel’s ability to preserve technological independence and maintain AI leadership.

Large-scale computing requires electricity and a great deal of it. Data centers and supercomputers consume enormous amounts of power and require cooling systems and appropriate physical infrastructure. Investing in artificial intelligence and supercomputers therefore requires suitable electricity generation and transmission infrastructure, land allocation, and the construction of data centers.

This challenge is no longer theoretical. The national plan published by the new Directorate in 2026 identified the need to address the growing demand for server farms and their connection to the electricity grid.

If Israel wants to become an AI powerhouse, it must also become an infrastructure powerhouse.

An effective AI system needs information. Israel is among the global leaders in this area and possesses enormous quantities of relevant data in healthcare, transportation, education, science, defense, and other government sectors. Some of this data can become a significant national resource and leverage.

Naturally, a delicate balance is required. Not every database can be opened. Citizens’ privacy, medical information, and sensitive security information must be protected. At the same time, safe and controlled mechanisms must be developed to allow researchers and companies to use appropriate information for research and development, while protecting individual privacy and the confidentiality of certain types of data, particularly defense-related information.

If computing is the engine of artificial intelligence, data is its fuel. Without it, we will not move forward.

One possible mistake would be for Israel to try to compete with the technology giants in every field. Israel does not necessarily need to build the world’s largest model, possess the largest number of processors in the world, and certainly does not need to manufacture all of them itself.

Israel must identify the areas in which it has comparative advantages and focus on them. There are fields in which Israel already has exceptional experience: software, cybersecurity, defense, intelligence, autonomous systems, chip design, medical technologies, and applications that combine the civilian and defense sectors.

Competitive advantage does not require the ability to do everything. Sometimes the greatest advantage lies in knowing where not to compete, and where to focus.

The committee also recommended strengthening advanced research at the national level and establishing a framework capable of connecting academia, government, the defense establishment, and industry.

The reason is clear. Private companies generally need to focus on products and markets, while academia can, and must, allow itself to think ten or twenty years ahead.

A country that wants to lead and remain at the technological frontier must simultaneously support both commercial innovation and long-term basic research.

The committee found that although Israel possesses the fundamental conditions required for success, we are currently behind. It is difficult, for example, to identify an Israeli-developed model that successfully competes with models produced in the United States by companies such as Anthropic, OpenAI, and Google; in China by companies such as Moonshot and DeepSeek; or in France by Mistral.

At the same time, a significant share of the human capital working in this field around the world has ties to Israel.

The IDF, particularly the Military Intelligence and the C4I and Cyber Defense Directorate, possesses large quantities of high-quality, diverse, and unique data suitable for training. With the right investment and direction, the talent developed within the IDF could generate substantial change within only a few years, provided these individuals are directed toward academia or appropriate positions within the IDF after receiving the necessary academic training.

Just as the IDF’s cyber capabilities helped fuel Israel’s and the West’s cybersecurity industry, AI experts trained academically for IDF service, through suitable elite programs, can help propel the Israeli ecosystem toward global leadership.

There is a fundamental difference, which not everyone understands, between cybersecurity experts and AI experts. A cybersecurity expert can be a talented 18-year-old hacker who sits in front of a computer and understands how to “break into” systems and identify vulnerabilities. That is not the case with AI.

A core AI expert, unlike an ordinary AI user, must receive appropriate academic training and should preferably hold at least a master’s degree, and ideally a doctorate. This field requires significant national-level investment in computing infrastructure, servers, and energy, as well as in human capital and research and development.

In recent years, Data Science has developed into a recognized profession within the IDF and is the field most closely related to AI research and development. In this area, the IDF trains dozens of researchers at the highest level each year through programs such as “Alonim” and “Bareket,” where they operate in a high-talent-density environment that enables excellence to flourish.

On the development side, the software industry is changing, and the number of junior-level positions worldwide is declining. This raises an important question: how can the labor market produce senior professionals in sufficient numbers and with the necessary knowledge and experience when it is becoming increasingly difficult for junior employees to enter the workforce and begin gaining experience?

The IDF is expected to continue playing a significant role in training junior professionals, functioning as a large-scale system for training, placement, and experience-building that relies on both mandatory-service soldiers and academic reserve program participants.

Training for the AI era is also necessary for the IDF’s core professional roles, commanders, combat soldiers, and intelligence personnel. Serving alongside AI agents, and particularly in multi-agent environments, especially during time-sensitive and critical operational processes, requires training, adaptation, and the development of appropriate doctrine, just as with the introduction of any new technology.

In my previous article, I emphasized the critical importance of using AI within the IDF and developing systems that assist in decision-making, but do not independently make decisions involving life and death.

Artificial intelligence also creates risks: privacy violations, misinformation, misuse of systems, information-security challenges, and the implications of deploying AI systems in critical fields.

Yet regulation that fails to keep pace with technological change can become an obstacle. A flexible approach is required: one that protects the public and national security without creating a situation in which Israeli innovation moves elsewhere because of excessive bureaucracy. Technology will change rapidly, and regulation will need to adapt accordingly.

The scale of investment recommended by the committee may appear very substantial: approximately NIS 25 billion over five years.

This may seem like a large amount, but given the importance of the field and the investments being made by competing countries, it is actually quite modest, smaller in scale than the investments being made by countries such as Saudi Arabia and the United Arab Emirates, not to mention the major global powers.

Nevertheless, the central question is not simply how much the government invests, but how it manages that money.

Every investment should be accompanied by measurable objectives: How many researchers will be trained? How many companies will grow? How much computing power will Israel have? Which infrastructure will be built? And what economic and national security benefits will result?

The Nagel Committee’s recommendations did not remain on paper, as has happened with many committees in the past, although not all the recommendations, particularly the proposed budget and the method for building the new Israeli supercomputer, have yet been fully implemented.

The 2026 national strategic plan that was published includes, among other things, a target supercomputer of up to 100,000 GPU's to be built through the private market, despite the committee having recommended a different approach. This target has not yet been funded.

The plan also includes promoting infrastructure for server farms, examining the establishment of an advanced semiconductor manufacturing facility, advancing an Israeli quantum computer, expanding artificial-intelligence studies, and training workers.

This means that the debate has shifted from whether Israel should invest in AI to the much more practical question of how to do so correctly and at what pace.

The Nagel Committee’s recommendations sought to change the way artificial intelligence is perceived in Israel. AI is not simply another high-tech sector that can be encouraged through a few grants. It is national infrastructure, comparable to electricity, communications, healthcare, science, and national security.

The question that will ultimately determine the success of this effort is not merely how many billions are allocated, but whether Israel can execute.

If we succeed in connecting human capital, academia, industry, the defense establishment, and national infrastructure, artificial intelligence can become a powerful engine of growth and a significant source of strength for Israel.

If we do not, there is a risk that Israel’s potential advantage will erode precisely in the field that is expected to shape the upcoming decades.