AI push stalls inside Israeli government despite national tech strength, report finds

“Artificial intelligence is not a future issue. It is already changing the way governments operate,” Englman said. He added that audit institutions must examine government preparations before the risks materialize, not after.

The findings said AI adoption in government must improve public service while protecting individual rights and public trust. Implementation should strengthen transparency, safeguard human rights, ensure responsible use of state resources, and make public services safer, more efficient, and more beneficial.

Although the government adopted the recommendations of the Nagel Committee in September 2025 under Government Decision 3375 and ordered the establishment of the National Artificial Intelligence Headquarters in the Prime Minister’s Office, Israel still had not approved a comprehensive long-term national AI plan as of the audit’s completion date.

The gap is especially significant because Israel’s starting point is unusually strong. Its technological capabilities, high-tech industry, research activity, and human capital place it in a favorable position in AI, but those advantages have not yet become a full government implementation plan.

The Office of the State Comptroller conducted what it described as the first comprehensive mapping of its kind of Israel’s public sector readiness to adopt and implement AI. The findings were based on responses from 70 leading public bodies, including most government ministries, statutory bodies, auxiliary units, hospitals, health funds, large municipalities, and other entities.

The questionnaire found broad managerial recognition of AI’s importance. In 77% of the bodies, management attached great or very great importance to integrating AI solutions; in 63%, there was a leading and coordinating figure in the field; and in 72%, employee training programs were already being operated.

The public bodies reported 144 AI projects across 47 entities. Of those projects, 42% supported the core activity of the ministry or body, and 34% were intended to improve service to citizens, indicating real potential to improve public-sector performance.

Despite the growth in AI activity, only 18% of the bodies reported that they had adopted a defined organizational strategy or policy for integrating AI.

The findings also showed that 34% of the bodies had not yet begun formulating a data strategy, while 41% operated without a formal data-governance framework. These gaps could limit the public sector’s ability to use AI responsibly, because AI systems depend on high-quality, accessible, secure, and well-governed data.

Of the reported AI projects, 68% were still in development or pilot stages, while only 32% had been implemented in practice. The findings said this shows that Israel’s public sector has not yet created the broad and mature infrastructure needed to move from isolated experiments to safe, effective, and measurable implementation.

Budget gaps are slowing AI adoption

The findings said 58% of participating public bodies had not been allocated a dedicated budget to promote AI projects during the years examined. Englman said the absence of organizational and budgetary infrastructure in most public bodies is delaying the development of public service.

About 80% of the bodies pointed to dedicated budgeting as the support most needed to accelerate AI adoption. Another 62% pointed to the need for training, while 58% said procurement mechanisms must be made more flexible.

“In order for Israel to realize its status as an ‘innovation nation’ within government bodies as well, we must face the gaps emerging from the field,” Englman said. “Now is the time to formulate a national master plan that will turn artificial intelligence tools into a lever for excellence in government service.”

The questionnaire found that 86% of participating public bodies do not have autonomous decision-making systems based on AI. The finding suggests that most public-sector AI use remains focused on support tools, service improvement, internal efficiency, or projects that have not yet reached full operational implementation.

The findings framed this within the broader need for responsible implementation, legal and ethical guidelines, information security, privacy protection, and benefit-measurement tools before AI systems are used more widely in public administration.

Israel has a policy for information sharing, but public bodies still face major obstacles, including prolonged approval processes, regulatory and bureaucratic limitations, enforcement gaps, information systems that do not interface with one another, and dependence on manual processes.

In the absence of an orderly and measurable government data strategy, it is difficult to turn government databases into reliable infrastructure for data reuse, advanced analysis, and responsible AI implementation in public service.

Responsible AI implementation must be built on high-quality data, proper data governance, information security and privacy, skilled human capital, and clear managerial responsibility.

The report said that human capital in the public service is critical to responsible AI adoption. Although Israel is strong in technological human capital and research, the government requires a complementary capability among civil servants, managers, regulators, procurement personnel, legal advisers, information-security personnel, and internal auditors.

These officials must be able to understand AI technology in depth, assess its risks, supervise external suppliers, and ensure that its use serves the public.

Localized training is not enough, according to the findings. A cross-cutting policy is needed to develop AI literacy and competence across the public service.

Responsible AI implementation could become a central lever for improving public-sector efficiency and strengthening service to citizens. However, Israel must shift from viewing technological innovation as a localized project to treating AI as a cross-government capability.

A comprehensive government framework is required, combining uniform policy, dedicated budgeting, secure data and cloud infrastructures, professional training, adapted procurement mechanisms, legal and ethical guidelines, and tools for measuring benefits.

Englman also spoke of the state’s obligations with AI. “It is our duty to ensure that the adoption of technology promotes high-quality and efficient public service, while protecting individual rights and public trust,” Englman said.

Separate reporting has also raised questions about how the new AI authority will interact with existing government bodies involved in technology, cyber policy, and national security.