Data privacy in Creatio AI
Creatio does not have its own LLM (large language model) and does not train the LLM used in Creatio AI on Customer data. Creatio employees do not have access to customer data or Customer dialogues with Creatio AI.
The LLM provider does not process customer data. LLM as software is hosted and operated on Microsoft Azure services in the USA (for the US customers) or Sweden (for EU and other regions) depending on your Creatio website location.
Creatio AI assists with compliance regarding user access permissions and restrictions.
Read more about the LLM used in Creatio AI here: Data, privacy, and security for Azure OpenAI Service.
Creatio AI uses Customer Data for communicating with the LLM. This enables it to get clear and useful answers when you ask Creatio AI something. We call that data context.
Context is the information provided to the LLM. It helps the LLM to understand what is expected of it and ensures the accuracy and relevance of its responses.
Creatio platform uses 3 types of data as context:
-
Page data. Creatio AI can request and utilize data from the page where you where are currently working.
-
Message history. This is your chat history with Creatio AI. It helps the LLM to understand the context of previous conversations, enabling it to better interpret your requests and stay within the scope of a single query as long as you need it.
-
Additional context from AI Skills and business processes. If your AI Skill has actions configured that use Customer Data as output parameters, Creatio AI can use this data to generate responses, trigger functions, etc.
Creatio AI does not have access to previous chat history if you have clear it or start a new conversation.
We strongly recommend using Creatio AI only for the issues related to your usage of Creatio services for the internal business purposes, as described in your agreement with Creatio.