5 ways to use AI to challenge executive assumptions
When instinct outranks evidence in the boardroom, here is what to do about it.
Stephanie Nivinskus is principal at Ragan’s Center for AI Strategy.
Conversations become awkward quickly when an executive defends a position on your company’s reputation, competitive standing or customer sentiment, and you know their claim rests on their instinct rather than anything verifiable.
As John Adams once said, “Facts are stubborn things.” But you know it’s not appropriate to make the conversation about who is right. So how do you close the distance between truth and fiction without the conversation going sideways?
AI can help comms leaders challenge executive assumptions without creating unnecessary friction. Here is what advisors to Ragan’s Center for AI Strategy had to offer:
Catherine Richards, founder of Expera Consulting, said you should “use AI to gather evidence from outside the company, then use your experience and judgment to determine whether it supports or challenges what leadership believes to be true.”
The move shifts the conversation from your interpretation to outside data — evidence that arrives without your fingerprints on the conclusion.
Michael Lamp, chief digital officer at HUNTER, said the approach is “a great case for agent building over basic use of existing/commercial large language models.”
He described a faster path than many expect: “Within a few steps, you can build a knowledge base of internal documents and vibe code your way to a private agent that uses a finite amount of information to deliver against a narrow brief.”
An agent scoped to a specific question produces a specific answer, and that specificity is what makes it useful in a leadership conversation.
Jonny Bentwood, global head of data and analytics at Golin Ketchum, named the same move in plainer terms: “Let the data do the disagreeing. When a CEO insists ‘everyone knows us for this reason,’ run the evidence: what the models say, what surfaces when customers ask, where the perception gap sits.”
Chris Gee, founder of Chris Gee Consulting, pushed the preparation earlier. Before the meeting, he said, “load your executive’s own memos, town halls and earnings calls. Then, build out key stakeholder personas and let AI argue their side back at you.”
The result: “You’ll either kill your own point in 20 minutes or walk in with their objections already answered.” The friction disappears when you have already had the argument before the room fills.
Anika Gakovic, Phyusion, reframed the use case as a governance tool rather than a tactical move. She said comms leaders can “facilitate ‘let’s run the AI model’ pre-mortems that generate ‘plausible set-back’ scenarios to prepare executive team responses.”
Where traditional strategic planning is bound by the assumptions leaders already hold, AI surfaces the threats they have not yet imagined, and makes the conversation collaborative rather than adversarial.
What to have ready before the next leadership meeting:
- Run the outside evidence first. Gather sentiment, coverage or model-generated answers on the claim in question before you walk in so you can support your position rather than assert it.
- Build the agent before you need it. Scope one internal knowledge base to one recurring brief. A narrow agent will produce a specific answer, and a specific answer is harder to dismiss than a general one.
- Pre-mortem the point you plan to make. Load relevant internal documents, build the opposing stakeholder persona and let AI play devil’s advocate. Walk in with the objections already resolved.
