Lam, Sean Greetings & Welcome to Microsoft Q&A forum!
Is there a method in which the model will restrict it's answers based on users' access and permissions to sensitive data?
Please note that the model itself does not have the capability to identify the user and restrict the information it gives based on their access and permissions to sensitive data.
You can try by implementing access controls and authentication mechanisms to ensure that only authorized users have access to sensitive data. Additionally, the model can be trained on a subset of the data that excludes sensitive information to prevent it from generating responses that include that information.
You can control access to the Azure OpenAI Service resource. Please see Role-based access control for Azure OpenAI Service for more details.
Also, see How to configure Azure OpenAI Service with managed identities and Restrict your Azure AD app to a set of users in an Azure AD tenant and let me know if that helps.