OpenAI releases healthcare AI policy blueprint

 OpenAI has released a healthcare policy blueprint calling for broader patient access to medical data, expanded clinician use of AI tools and updated regulatory pathways for AI-supported care.

The report, “Keeping Patients First: A Blueprint for AI in U.S. Healthcare,” was published in April and argues AI can help address barriers related to healthcare access, affordability and administrative burden.

Here are eight key findings from the report:

1. The blueprint centers on three broad themes: patient-directed data portability, clinician-supervised AI deployment and modernization of healthcare regulation.
   
   
2. OpenAI said patients should have expanded rights to access and aggregate their healthcare information across systems, including laboratory records, pharmacy data and patient-generated information such as wearable device data. The report also called for broader interoperability requirements and stronger enforcement of federal information-blocking rules.
   
   
3. The report states AI systems should not replace licensed clinicians or present themselves as healthcare professionals.
   
   
4. At the same time, OpenAI said clinicians should be permitted to use AI for administrative and workflow support functions, including transcription, documentation and summarization tasks. The report argues such uses could reduce physician burnout and increase time spent on direct patient care.
   
   
5. The document also calls for new “regulatory sandboxes” that would allow health systems, clinicians and AI developers to test AI-supported care models under defined safeguards and oversight structures.
   
   
6. In addition, OpenAI urged the FDA and HHS to clarify regulatory pathways for AI-enabled medical software and establish review processes for higher-risk systems intended to support diagnosis or treatment decisions.
   
   
7. The report includes several case studies describing existing AI deployments, including one example of AdventHealth, based in Altamonte Springs, Fla., using ChatGPT for Healthcare to assist with post-discharge outreach documentation. According to the report, the pilot reduced documentation time and increased outreach capacity for care coordination specialists.
   
   
8. The report concludes broader adoption of AI in healthcare should prioritize patient access, clinician oversight and equitable distribution of benefits.

https://www.beckershospitalreview.com/healthcare-information-technology/ai/openai-releases-healthcare-ai-policy-blueprint/