Imposter Syndrome Using AI for Presentations: When You Feel You Are Cheating

Ines is a director of clinical operations in a mid-size pharmaceutical company. She had been using Copilot for three weeks before the feeling caught up with her. The feeling arrived during a steering committee meeting, mid-sentence, while she was presenting a deck she had drafted with AI assistance. She was making a strong point about supply chain resilience when an internal voice cut in: “You did not write this. You should not be presenting this. If they ask you something the deck does not cover, they will see you do not actually know it.”

The voice was loud enough that she lost her place for half a second. The committee did not notice. She recovered. The presentation went well. But the feeling stayed with her for the rest of the day and crystallised that evening into a question she put to a colleague over dinner: “Am I cheating? Should I just write the decks myself like I used to?” Her colleague, who had been using Copilot since launch, said something useful: “If you wrote the prompt and you read the output and you decided what to keep and what to change, you wrote the deck. The keyboard is not where the work happens.”

That sentence is technically correct, and it does not always land in the moment because imposter syndrome is not technically responsive. The cheating feeling has its own logic, and arguing with it head-on rarely works. What does work is understanding why the feeling shows up specifically with AI — and then applying three reframes that change the underlying perception, not just the surface argument.

Why the cheating feeling shows up

Imposter syndrome activates when there is a perceived gap between what others believe you contributed and what you privately know you contributed. AI use opens that gap by design. The audience sees a polished deck. You know that some of the structure came from a model. The two pictures do not match in your head, and the mismatch reads as deception.

The feeling intensifies if your professional identity is tied to “I produce my own work”. Many senior leaders built their careers on visible production — writing the strategy memo, building the financial model, drafting the board paper themselves. AI changes the labour mix. You still own the output, but the labour is distributed differently. The labour distribution change feels like an identity threat, even when the output quality is equal or higher.

It also intensifies in environments where AI use is technically allowed but socially ambiguous. If your employer has not explicitly endorsed AI for presentation work, but has not explicitly forbidden it either, you are operating in a grey zone. The grey zone amplifies imposter feelings because there is no external validation that what you are doing is acceptable. Your nervous system fills the validation vacuum with the worst-case interpretation: that you are doing something you would not want to admit to.

Visible drafting versus invisible judgement

The cleanest way to understand what is actually happening is to separate the visible work from the invisible work. The visible work in a deck is the typing, the layout, the wording of bullets, the choice of charts. The invisible work is the prior thinking — what to include, what to leave out, what the argument should be, which evidence carries weight, how the audience will react, where the political risk lies, what the closing decision needs to be.

For a senior-level presentation, the invisible work is roughly eighty per cent of the value. Anyone with passable Copilot skills can produce a polished thirty-slide deck on any topic in twenty minutes. Almost no one can produce a deck that lands with a specific board on a specific decision in a specific organisational moment without the invisible work that comes from years of internal context.

When you use AI for the visible work, you are outsourcing the part that has the lowest unit value of your time. You retain the invisible work — the editorial judgement that decides which AI output to keep, which to rewrite, which to cut, which to anchor with internal evidence the model could not have known. This is the work the audience cannot see, and it is also the work that your imposter voice is failing to credit. The voice notices that you typed less. It does not notice that you decided more.

Reframe one: the typing is not the work

The first reframe is to separate effort from value. There is a deeply ingrained association between visible effort and earned credit, particularly in cultures where being seen to work hard is part of the professional identity. AI breaks that association by making the visible effort smaller while leaving the cognitive load roughly constant.

The reframe is simple to state and harder to internalise: the typing is not the work. The work is the judgement applied to what gets typed. A surgeon’s value is not in the physical incision — it is in knowing where, how deep, and when to stop. The incision is the visible part. The training and judgement underneath are the invisible part. AI makes the executive presentation analogous to the surgical analogy. The model does the incision. You do the judgement.

This reframe lands harder when you can name a specific decision you made on the most recent AI-assisted deck that the model could not have made. “I cut the section on European expansion because I knew the chair would push back on the timing — the model did not know that.” “I rewrote the headline on slide eleven because the original was technically correct but politically tone-deaf for our CFO — the model did not know that.” Naming the specific decisions that required your judgement is the most direct route to dissolving the cheating feeling. The decisions are real. They are the work.

Reframe two: AI is a tool, not a co-author

The second reframe targets the way the imposter voice tends to anthropomorphise AI. The voice often phrases the concern as “the AI wrote this, not me” — which assigns agency to the model. The model has no agency. It cannot decide what to write. It can only produce probabilistic next-tokens based on the prompt you supplied and the editorial decisions you made along the way.

The framing that helps is to compare AI to other tools you do not feel imposter syndrome about. You do not feel guilty using Excel to calculate a forecast you could have done by hand. You do not feel guilty using PowerPoint instead of drawing slides on acetate. You do not feel guilty using a spell-checker. The reason is that those tools are clearly tools — they execute under your direction, they have no agency, they do not “co-author” the output.

AI feels different because it produces something that looks like prose, and prose feels like authored content. But the AI is no more an author of your deck than Excel is the author of your forecast. It is a tool that executes your direction. The difference between a Copilot draft and an Excel formula is purely surface-level — both are deterministic outputs of inputs you supplied. The structured workflows that produce executive output reinforce this — the agent is following your instruction set, not writing the deck.

Reframe three: the question your imposter voice is really asking

The third reframe goes one layer deeper. The “am I cheating” question is rarely the actual question underneath. When senior professionals dig into what the imposter voice is genuinely worried about, the underlying question usually turns out to be one of three things, and each one has a different response.

The first underlying question is “if they ask me something off the slides, will I look foolish?” This is a competence question, not an authorship question. The answer is not to abandon AI — it is to do the depth work that prepares you to answer questions beyond the deck content. The deck is one slice of your knowledge. AI helped you produce the slice. Your years of context are what handle the questions. Use the time AI saves you to deepen your audience preparation, not to do less work overall.

The second underlying question is “if they find out I used AI, will they think less of my contribution?” This is a social-acceptance question. The honest answer is that some audiences will, particularly in environments that are still adjusting to AI norms. The right response is not concealment, which feeds the imposter voice. The right response is matter-of-fact disclosure when asked, framed around the editorial judgement that produced the final output: “Yes, I used Copilot to draft the structure; the analysis and the recommendation are mine. The AI saved me about three hours.”

The third underlying question is “if AI can do this, what am I actually contributing?” This is an identity question, and it deserves a serious answer rather than a deflection. AI cannot do the invisible work — the situational awareness, the political read, the executive context, the judgement that comes from having been in the room before. Those are your contribution. AI use highlights this contribution by stripping away the typing that used to obscure it. If you find your contribution unclear after AI strips the typing away, that is useful information about where to focus your professional development. The right response is to invest in the parts of your work AI cannot do, not to retreat from AI use to preserve the visible parts it can.

The productive caution worth keeping

None of these reframes are about silencing all hesitation around AI use. There is a productive caution underneath the imposter feeling that is worth preserving — the caution that prompts you to verify numbers the AI generated, to check the source of claims, to read the deck aloud against the audience’s likely reaction, to take responsibility for what reaches the room. That caution is the editorial judgement at work. Keep it. It is the difference between AI-assisted senior output and AI-flavoured generic output.

The reframes target the unproductive part of the feeling — the part that says you are not entitled to present material because you used a tool to draft it. That part is wrong, and feeding it makes you a worse presenter, not a more honest one. Concealing AI use because the imposter voice told you to leads to evasive answers when audiences ask direct questions, which damages credibility more than the AI use itself ever would.

The senior professionals who handle this transition cleanly tend to land on a stable framing: AI is a tool I use to do my work faster; the work itself — the judgement, the decisions, the editorial pass — is mine; if asked, I will say so plainly; if not asked, I will not perform a confession that is not required. The editorial pass is what makes the difference between AI output that lands and AI output that gets pushed back. That pass is yours. The cheating voice is misreading the labour. Do not reorganise your career around its mistake.

FAQ

Should I tell people I used AI to draft the deck?

If you are asked directly, yes. Honesty handles the question once and removes the imposter loop entirely. If you are not asked, you do not owe a proactive disclosure unless your organisation requires one. Performing a confession that was not requested often draws more attention to AI use than a matter-of-fact answer would. The framing that works in either case is “I used AI to draft the structure; the analysis and recommendation are mine” — which credits both the tool and the judgement honestly.

Why does the cheating feeling get worse the better the AI gets?

Because the gap between visible AI contribution and invisible human judgement gets larger as the model improves. Earlier AI tools produced obviously rough output that you visibly had to fix; the editorial work was visible because the gaps were visible. Better models produce smoother output that needs subtler editorial work; the gaps are no longer visible to you, even though they are still there. The judgement work has not disappeared — it has just stopped being noticeable. The reframe is to deliberately track the editorial decisions you are still making, even when they feel small.

Is imposter syndrome about AI different from regular imposter syndrome?

It has the same underlying mechanism — a perceived gap between contribution and credit — but a different trigger. Regular imposter syndrome is triggered by promotion, scope expansion, or visibility increases. AI-related imposter syndrome is triggered by the labour distribution change. The mechanism is the same; the trigger is new. The same techniques that help with regular imposter syndrome — naming specific contributions, reality-testing the worst-case interpretation, talking to peers — also help here. The first reframe in this article is the AI-specific addition.

What if my anxiety about using AI is severe enough to disrupt my presentation performance?

If the cheating feeling intensifies during the presentation itself rather than dissolving with the reframes, the underlying issue is performance anxiety more than imposter syndrome about AI specifically. The AI use is the trigger but not the cause. Practical techniques for in-the-moment anxiety — controlled breathing, the structured pause, the recovery sentence — work the same way regardless of whether AI was involved in producing the deck. The deck is yours to present once you are in the room. The earlier the anxiety pattern is addressed, the less it will surface in subsequent presentations.

The Winning Edge — Thursday newsletter

Every Thursday, The Winning Edge delivers one structural insight for executives presenting to boards, investment committees, and senior stakeholders. No general tips. No motivational framing. One specific technique, one executive scenario, one action. Subscribe to The Winning Edge →

Not ready for the full programme? Start here instead: download the free Pyramid Principle Template — the framework that gives your editorial judgement a structural target to edit toward.

Next step: name three specific editorial decisions you made on the last AI-assisted deck you produced. Write them down. Re-read them when the cheating voice next surfaces. The decisions are real. The voice is misreading them.

Related reading: The Copilot Agent Mode workflow that makes editorial judgement the senior contribution.