Should Rovo challenge users instead of always trying to help them?

[AGASTYA SHARMA]

Today's AI assistants are designed to answer questions.

But should enterprise AI also question assumptions?

For example:

"This requirement conflicts with an existing Jira Epic."

or

"This architecture contradicts your documentation."

Would a "critical thinking mode" make Rovo more valuable, or would it become intrusive?

[Rebekka Heilmann _viadee_ · Accepted solution]

You can prompt Rovo to find conflicts and contradictions. You could also set this up in Rovo's memories.

Note: This is not really a Question but rather a discussion, best suited for the discussion forum: Atlassian AI & Rovo - Atlassian Community

[Dr Valeri Colon _Connect Centric_]

@AGASTYA SHARMA you can already configure this behavior in a custom Agent—for example, instruct it to challenge assumptions, identify conflicts with Jira/Confluence evidence, and clearly flag missing information. I’d put that in the Agent’s Behavior/Instructions rather than Memory, since Memory is personal to each user. So a “challenge mode” is possible today, even though it isn’t a named Rovo mode.