Use of AI Tools

2023.3

Language Learning models (LLM), like ChatGPT, are powerful tools for generating text and responding to prompts. These models are used in a variety of applications, including chatbots, language translation, and content creation. This policy addendum outlines the guidelines for the use of LLMs within LifeOmic.

In summary, LifeOmic employees may use tools such as ChatGPT for small tasks, but sensitive workloads or production requests should never be handled by uncontrolled AI models. The models should be treated like any other non-approved third party or vendor.

LLMs owned, hosted, and maintained by LifeOmic are exempt from the details of this policy.

Purpose of Use#

LLMs should only be used for non-regulated or confidential projects. Some examples:

• Document template generation
• Marketing language review

Data Security#

LLMs may NEVER process sensitive or confidential data, including patient information.

Quality Assurance#

LLMs are not infallible and will make errors or generate inappropriate responses. LifeOmic does not use outputs from LLMs in any published research, operational decisions, or coding guidance.

Compliance#

All use of LLMs must comply with local and federal regulations. For instance, employees must not disclose any confidential information or violate any data protection laws while using a LLM.