Analysis | How Israel is Building Its Digital Sovereignty in the AI Era

While the global high-tech market wakes up almost every morning to a new trend in the world of artificial intelligence, the war in Israel has turned the concept of data sovereignty from a technological term into an existential necessity.

In days when the country is under the threat of missiles targeting critical infrastructure, alongside scenarios of blackout, a deep and strategic process is taking place beneath the surface: the construction of national cloud infrastructure. The goal is clear. It is not merely technological progress, but ensuring the state’s functional continuity for decades to come, from the establishment of resilient data centers to the consolidation of national planning data.

Recent incidents of cloud infrastructure outages in various countries over the past year serve as a warning sign for every organization in Israel. They highlight the gap between the perception of cloud as inherently resilient and the strategic reality of regional centralization, dependence on physical infrastructure, and the fragility of supply chains, which together can lead to large-scale outages.

Independence that Promotes Efficiency

A need that has sharpened in recent years is the requirement for a sovereign cloud. The State of Israel is learning that it cannot rely on public cloud concepts hosted in foreign countries or owned by foreign companies subject only to their domestic laws, as these may become unavailable. Cloud is a powerful platform, but for a country under threat, the lesson is clear: we must secure the future of our data on our own territory through data centers and hybrid cloud infrastructures.

The transition to architectures combining public cloud with local and secure instances (hybrid cloud incorporating confidential computing), along with the implementation of secure remote access solutions, is not a luxury but a condition for ensuring service continuity during emergencies. This is intended to prevent situations where essential services are shut down due to the state, ethical, or commercial considerations of external providers. Alongside this, there is a growing need for data consolidation within secure environments controlled by the government and its cyber units.

Projects carried out in the civilian sector, such as “Agma Data,” established for the Planning Administration, demonstrate how consolidating data previously dispersed across government ministries and local authorities into a single repository can shorten national processes from months to days. 

This shift aligns with the broader trend seen in 2025 and expected to continue in the coming years, where big data and AI are no longer experimental tools but strategic instruments. Despite budget constraints, organizations now understand that real value lies in building a robust infrastructure that combines public cloud systems while enabling organizational knowledge to be accessed intelligently and securely. The ability to transform “fragmented and dispersed” data into an accessible operational resource will determine who leads in the coming decade.

Such projects, based on the integration of advanced cloud infrastructure and big data, serve as the “fuel” powering generative and national AI models. The correct step forward is to lead similar processes, including the planning and establishment of AI-optimized server farms and national infrastructure, with the understanding that a data center is not just an empty physical space, but a complex system requiring automation and intelligent management.

What Is Still Holding Us Back?

Despite progress and development momentum, the path is not without challenges. The main reason is state budget limitations and complex regulation that must catch up with the pace of innovation. Another challenge is the need for security and regulatory approval of Confidential Computing solutions, which allow the use of sensitive data securely within public and hybrid cloud infrastructures, alongside less sensitive data.

In the context of AI, Israel must continue investing in national infrastructure to maintain momentum. With the development of national language models, the goal is to lead in language and deep applications, as is already happening today in medicine, research, and national infrastructure domains. 

The establishment of a national AI administration is a very important step in the right direction, but its success depends on the ability to implement strategy across all relevant verticals awaiting transformation (digital government, education, healthcare, transportation, agriculture, defense, and more), recognize AI farms as a national asset, and make AI resources accessible both to research bodies and to the public, in order to prevent digital inequality in Israel.

Anatoly Kushnir is the CTO and Head of ICT Division at Commit.