Health Ministry Blocks AI Tools in State Hospitals Over Patient Privacy Fears

Israel’s Health Ministry has blocked access to major AI tools, including ChatGPT, Claude and Gemini, on all computers in state-run hospitals. The move, carried out without advance notice, was reported this week and surprised staff at hospitals such as Sheba Medical Center and Assaf Harofeh Medical Center.

The ministry said the decision is intended to protect patient privacy, prevent leaks of sensitive medical information and reduce cyber risks to public medical institutions. From now on, hospital staff can use those AI tools only on their own personal mobile devices, not on hospital computers connected to institutional networks.

The restriction is a sharp escalation from guidance issued in March 2026, when the ministry told public health organizations to minimize use of external AI systems. That earlier warning focused on the threat of cyberattacks against Israeli hospitals and instructed employees not to enter organizational data, medical information, personnel details or code into outside AI platforms.

In its formal response, the ministry said that “protecting patient privacy, medical confidentiality and information security are top priorities” and that it sees AI as an important engine for innovation in healthcare. It added that free, public external AI tools create cyber risks and could expose sensitive medical data. The ministry said it is also working on dedicated protective solutions that would eventually allow wider AI use, even on networked hospital computers, while maintaining high privacy and cybersecurity standards.