Health system-wide access to generative artificial intelligence: the New York University Langone Health experience

Area of Science:

• Healthcare Technology
• Artificial Intelligence in Medicine
• Digital Health Transformation

Background:

• Academic health centers are exploring generative artificial intelligence (GenAI) to enhance operations.
• Implementing secure and private GenAI instances is critical for sensitive healthcare environments.

Purpose of the Study:

• To evaluate the adoption, usage patterns, and perceived impact of a private GenAI application within a large academic health center.
• To understand employee interactions with GenAI and its influence on perceived skills and work performance.

Main Methods:

• Establishment of a secure, private Azure OpenAI service (GenAI Studio) at New York University Langone Health (NYULH).
• Widespread employee access granted, with usage monitoring and user surveys conducted over six months.

Main Results:

• Over 1007 individuals applied for access, with significant utilization in research and clinical departments.
• Users reported ease of use, preparedness, and willingness to recommend the GenAI studio.
• GenAI was applied to diverse tasks including writing, editing, summarizing, data analysis, and idea generation.
• Identified challenges included workforce prompt engineering education and technical limitations (token/API).

Conclusions:

• High interest and extensive use of GenAI in healthcare settings were demonstrated.
• GenAI serves as a valuable tool for augmenting employee skills and daily tasks.
• Effective workforce education and guidance are essential for successful system-wide GenAI implementation.