I think that "communication pattern" is inaccurate information. But maybe that points to the real issue.

Do we agree that harmful inaccurate information should be screened for and eliminated?

That leads to a follow-up question: how do you screen for inaccurate information at scale?