Health Ministry blocks AI access on government hospital networks over cyber fears

The Health Ministry has blocked access to artificial intelligence tools on all computers in government hospitals, in a move aimed at strengthening patient privacy and reducing the risk of leaks of sensitive medical information and cyberattacks on public hospitals in Israel. N12 reported the step on Friday morning.

Staff at government hospitals, including Sheba Medical Center and Assaf Harofeh, discovered they could no longer reach commonly used AI systems through the hospitals' corporate networks. The blocked services include ChatGPT, Gemini and Claude. Medical teams who still want to use these tools will now have to do so on their personal devices, not on hospital computers.

The decision tightens guidance the ministry issued about four months ago, when it warned all public health organizations in Israel to limit the use of external AI tools as much as possible. That warning came amid concern that cyberattacks against Israel could also target public hospitals. Employees were told not to enter organizational information, medical data, personnel details or software code into outside AI systems because of the risk of exposing sensitive material.

In its response, the ministry said, "Protecting patient privacy, medical confidentiality and information security are top priorities whenever technologies are introduced." It added that AI is an important engine for innovation in health care, but that public, free and external AI tools carry cyber risks and could expose sensitive medical information. The ministry also said that in March 2026 it issued a directive requiring all health organizations to stop using such tools on corporate networks and move only to secure systems, while it works on protective solutions that would allow broader AI use, including on networked hospital computers, under strict privacy and cybersecurity standards.