What to know about using AI at USA Health
Although AI will play an increasingly important role at USA Health, it is critical that we all remain diligent about not uploading our patients’ protected health information (PHI) to unapproved AI platforms, including ChatGPT, Claude and Gemini.
The following is a message from Natalie Fox, DNP, chief executive officer, and Tyler Whetstine, chief information officer.
At USA Health, we are excited about the ways that artificial intelligence (AI) will allow us to enhance patient care. AI tools already available to our team members include Ambient AI Agents for Oracle Health and Athena Health, Illuminate for our lung nodule Incidental Finding program, as well as others. We look forward to continuing to expand the availability of AI tools for our employees in the near future, and will focus our efforts and investments on tools that enhance our patient experience, our caregiver experience, or help to create operational efficiencies.
Although AI will play an increasingly important role at USA Health, it is critical that we all remain diligent about not uploading our patients’ protected health information (PHI) to unapproved AI platforms, including ChatGPT, Claude and Gemini. Our IT team has implemented mechanisms to block information that appears to be PHI from being uploaded to these services from USA Health owned machines, but the responsibility to protect our patients falls on each of us.
As with all digital services, USA Health must have a signed Business Associate Agreement (BAA) in place with an AI service provider before employees are allowed to upload PHI. No matter what promises a service makes about encrypting or deleting data, uploading protected information to a service without a BAA violates our compliance requirements.
To address questions about the use of AI and reiterate the importance of securing patient information, we have approved a new policy specifically about the use of generative AI, developed guidelines for our team members, and put together a list of approved services:
New USA Health Generative AI Use Policy
We also have edited existing policies to add language specifically addressing the use of AI:
Cell Phones and Electronic Devices
Access to Medical Records for Research Purposes
Assigned Security Responsibility
Integrity (Data Authentication)
Safeguarding Protected Health Information
Thank you for your ongoing efforts to protect our patients and USA Health while utilizing AI and technologies that will continue to transform the health and care of the entire Gulf Coast.
