Generative AI speeds up the preparation of emails, briefing materials and presentations, but it creates a new management risk: a gap between the quality of the text and the depth of understanding of the person signing it. A leader can present a perfectly worded proposal and discover at the first follow-up question that they cannot explain its assumptions. AI should therefore help with formulation, not replace the manager’s own understanding of the decision.
The quality of management communication was long constrained by how well a person could structure and formulate an idea. Generative AI substantially lowers that barrier. From a few notes, it can create a persuasive email, argument or meeting brief. But this is also where a new risk appears: the output can be much more persuasive than the actual understanding of the person using it.
Wade Foster of Zapier describes this as an understanding gap. A leader delegates too much of the intellectual work and only discovers during follow-up questions that they do not know the underlying material well enough to support their own message.
For a manager, this is not merely a reputational problem. If AI creates a strategic recommendation and the person does not examine its assumptions, they may effectively approve an argument they could neither construct nor defend themselves. The linguistic quality of the output can create an impression of greater certainty.
It is therefore useful to distinguish between delegating formulation and delegating judgment. AI can help shorten a text, organize arguments, identify weak points or prepare alternative explanations. Responsibility for whether the inputs are correct and the conclusion makes sense must remain with the person.
A simple test is to defend the proposal without the tool. Before an important meeting, try to explain three things without opening the document: what assumptions the recommendation uses, what could change the conclusion and why the main alternative was rejected. If you do not know the answers, the text is not ready for your signature regardless of its stylistic quality.
The second test is the source trail. A number, comparison or high-impact claim must have a traceable origin. AI can sound persuasive even when it combines data that do not actually belong together.
The third rule concerns meetings. A leader should not present material whose logic they cannot explain without reading the prepared text. Follow-up questions are not an inconvenient obstacle; they are a check of genuine understanding.
Generative AI can therefore improve the quality of communication while making insufficient understanding less visible. Organizations need to assess not only the quality of the document but also the author’s ability to explain, defend and revise the decision when new facts emerge.
KEY TERMS
- Understanding gap: The difference between the quality of an AI-generated output and the user’s actual understanding of it.
- Delegating formulation: Using AI to improve structure and language while retaining human judgment.
- Delegating judgment: Leaving substantive evaluation and conclusions to the system without sufficient human control.
- Source trail: The ability to trace the origin of facts and assumptions used in a decision.