Health03:00 · Jul 30

Sheba Medical Center Partners with OpenAI to Integrate AI Using Israeli Patient Data

Calcalist
Translated & summarized from Calcalist by baba
The story · English

Sheba Medical Center in Israel has signed a strategic collaboration with OpenAI, granting its medical teams access to OpenAI's specialized medical AI models, including ChatGPT Health. This partnership marks a significant milestone in the integration of artificial intelligence within Israel's healthcare system, comparable in importance to the country's earlier vaccine agreement with Pfizer during the COVID-19 pandemic. Israel's centralized, digital health system with comprehensive patient data accumulated over decades is a unique asset attracting global tech and pharmaceutical companies.

The agreement allows OpenAI to incorporate Sheba's clinical protocols, treatment pathways, and policy documents directly into its AI models, potentially enabling tailored medical recommendations suited to Israel's population and healthcare environment. For Sheba, which operates the ARC innovation center, early access to these AI tools aims to assist medical staff in decision-making, validating clinical recommendations, and reducing administrative burdens such as medical documentation and prior authorization requests. This could lead to shorter patient wait times, improved continuity of care, and more personalized treatment.

The collaboration also raises critical questions about data ownership, patient privacy, and cybersecurity. While Israeli law recognizes patients as the primary owners of their medical data, the use of anonymized information by commercial companies like OpenAI without explicit patient consent remains controversial. Past criticism of data sharing during the Pfizer vaccine deal highlights ongoing concerns about transparency and confidentiality. Additionally, the Israeli Ministry of Health has warned about increased cybersecurity risks from public AI tools and has restricted access to external AI on hospital computers, though many clinicians continue using such tools on personal devices, sometimes sharing sensitive information via unsecured channels.

Beyond privacy and security, there are clinical risks associated with AI "hallucinations," where AI models may produce inaccurate or false information, potentially leading to misdiagnoses or harmful treatment decisions if relied upon without proper oversight. The Ministry of Health is working on regulatory frameworks to balance the benefits of AI in healthcare with patient rights and data protection. This partnership between Sheba and OpenAI could herald a new era in Israeli medicine, but it underscores the challenge of leveraging AI advancements while safeguarding patient trust and privacy in one of Israel's most strategic resources: its citizens' medical data.

Read the original at Calcalist
Open the live terminal