Before a quarterly business review, a rep asks Copilot to generate a summary of an enterprise opportunity. The summary looks polished and references recent logged calls and emails, but omits a significant pricing concession the customer raised verbally last week during a hallway conversation at a trade show that the rep never entered into Dynamics 365 Sales. What is the most accurate way to characterize this situation?

The accurate characterization is that Copilot generates its summary from the activity data that actually exists in Dynamics 365 Sales, so an unlogged hallway conversation simply is not part of what it can summarize; the fix is for the rep to record that conversation as a note before treating the summary as complete. The claim that the summary is inaccurate proof that Copilot is unreliable for opportunity research and should not be used before executive reviews is incorrect because it overstates the problem as a reliability failure of the tool, when in fact the tool performed exactly as designed on the data available to it; the gap is a data-entry gap, not a defect in the summary. The claim that Copilot should be reconfigured to automatically capture verbal conversations that were never logged is incorrect because there is no capability, nor is one intended, for Copilot to capture spontaneous offline conversations that were never logged through email, calls, or notes; summaries can only draw on connected, recorded activity. The claim that the gap indicates a licensing issue is incorrect because the missing information has nothing to do with licensing tier; it is missing because the rep never entered it, and no tier of Dynamics 365 Sales would surface a conversation that was never recorded anywhere in the system. The rep's responsibility is to close that data gap themselves.