Hi Atlassian,
We’re currently trialling Rovo and had a question about content controls.
Tools such as ChatGPT or Gemini restrict what users can ask based on policies defined by the provider. While we appreciate that Rovo already respects user permissions, we’re interested in whether it’s possible to additionally restrict certain keywords or phrases.
For example, the ability to block or limit responses when sensitive terms are queried, with the option to allow exceptions for specific user groups. Relying solely on page-level permissions requires a high level of trust that all sensitive content has been correctly restricted, and it doesn’t account for user error. As a result, there’s a risk that sensitive information could be surfaced through Rovo queries.
This type of control would help reduce the risk of data leakage while still enabling appropriate access where needed.