This article provides detail on the Microsoft Azure OpenAI architecture to provide clarity on how secure using Sofia GPT really is.
In short, all components of Sofia GPT (the AI engine, data, etc.) are contained within OnePlan's Azure subscription. This means that using Sofia GPT is no more or less secure than using OnePlan in general.
The AI model behind Sofia GPT is very secure. Sofia GPT is built on Azure OpenAI, which is a Microsoft/Azure native variant of the infamous OpenAI GPT AI model.
How this works is that Azure OpenAI hosts the AI engine deployment (eg. GPT-3/3.5/4/etc) inside our OnePlan Azure tenant, which provides this AI as a service. The AI engine is deployed inside of the OnePlan Azure tenant and is not deployed into cross/multi-tenant. Also, it is an independent service that only inherits the Azure security layer and does not compromise any other services in Azure.
So, using Sofia GPT has the same level of security as using OnePlan itself.