Atlassian Intelligence AI Security and Model Training

[Aaron Davis]

Hi! I'm looking into Atlassian Intelligence for Confluence to summarize our content and make content searching more intuitive for users.

Our company has certain regulations for the use of AI and I'm wondering if by enabling Atlassian Intelligence we are subjecting proprietary data to be used to train an AI model. Does anyone have any insight on the security of proprietary data stored in Confluence and if there are concerns of enabling Atlassian Intelligence? 

[Sandeep Nagaraja]

You have to check with Atlassian Team on what information is stored when AI is used.  

If Atlassian uses third party AI tools like OpenAI (ChatGPT) then you have to look into Terms and Condition on Data Storage of OpenAI with Atlassian. 

[Sascha Masanneck]

Hi Aaron,

Atlassian states: 

We do not use your input or output for any other purpose. Our partner, OpenAI, neither retains your inputs and outputs nor uses your inputs and outputs to improve its services.

Source: https://www.atlassian.com/trust/atlassian-intelligence

[kc]

If there's anywhere we can get info or better still a list of exactly what information is being send to Open AI for processing?

[Jackman, Emily]

I'm also interested in a more detailed list defining what type of data is being shared with 3rd parties within the use of Atlassian Intelligence.