A Quick Note on Compliant AI Use

Before you try this, remember: never put proprietary company information into an LLM unless your compliance team specifically allows it.

Most compliance teams are fine with using publicly available information (like a published paper), but always follow your company’s compliance policies first.

The Hack: Use AI to Anticipate Off-Topic Questions

Before your next KOL meeting, take the publication you’ll be talking and feed it into AI. Then use this prompt:

“You are an experienced senior MSL presenting this paper to a KOL. Give me 5 random, off-topic questions that could derail me in this meeting.”

I put in my favorite random dermatology clinical trial paper on dupilumab. Here’s an example of what comes back after using the prompt above:

1. Competitive Landscape Curveball
   “How does this compare to nemolizumab or other IL-31 inhibitors in terms of onset of action and QoL improvement?”
   Risk: drags you into cross-trial comparisons — a compliance minefield if you’re not authorized to discuss competitors.

2. Safety Deep Dive
   “You mentioned a 3.9% rate of conjunctivitis — do you have any data on long-term ocular safety or mechanisms behind this?”
   Risk: pushes you beyond the scope of the publication, into mechanistic details or unpublished data.

3. Real-World Evidence Challenge
   “These are clinical trial patients — do you have any real-world data showing similar efficacy in more complex patients, like those with multiple comorbidities?”
   Risk: debate on trial design and external validity, not covered in the paper.

4. Mechanistic Rabbit Hole
   “If IL-4Ra is blocking signaling on itch sensory neurons, why doesn’t dupilumab eliminate itch completely in all patients?”
   Risk: a nuanced mechanistic discussion that’s hard to simplify without losing clarity.

5. Off-Label/Cost Controversy
   “What about patients with mild PN who aren’t eligible for systemic therapy? Are you seeing off-label use? And how are payers reacting?”
   Risk: forces you into cost, access, and off-label policy issues — all high compliance risk.

Bonus AI Tip: Ask for 5 More

Don’t stop at one round. Run the prompt again and ask for 5 more curveballs. Often the second or third set uncovers entirely new angles. Get more tips on using AI with my free guide 

What to Do With the Questions

Here’s where the real preparation happens. Once you have your list:

1. Practice answering them out loud
   Say the answer as you would in front of a KOL. Notice if you ramble, get defensive, or sound unsure. Adjust until you’re clear and confident.

2. Check for existing backup material
   If the question is likely and data exists, check whether there’s an approved backup slide or resource you can reference. Knowing this in advance prevents scrambling in the moment.

3. Practice bringing it back on topic
   Some questions will take you off into the weeds. Prepare bridging phrases to steer the conversation back to the paper, for example:

• “That’s a related point but what this study really highlights is…”

• “That’s beyond the scope of this publication, but what we can say with confidence is…”

4. Plan your scenarios
   Some questions you’ll answer directly, others you’ll pivot or defer. The key is to plan which strategy fits each question so you aren’t deciding in the moment.

Why This Works for MSLs

Running this AI exercise helps MSLs:

• Anticipate derailers before they happen.

• Practice answering out loud so you sound confident under pressure.

• Strengthen executive presence with pivots and backup slides.

• Build trust with KOLs by staying composed even when pushed off-topic.

It’s like running a fire drill for your meetings: you hope you don’t need it, but if you do, you’re ready.

The Takeaway: Smart MSLs never skip this AI hack

Upload the publication. Run the prompt to identify off-topic questions. Ask for 5 more. Practice answering them. Prep your pivots and slides.

Because the difference between being flustered and being confident often comes down to how you handle the Q&A. Nothing is more rattling than getting cornered in a meeting. When you freeze under a tough KOL question, the room feels it. When you stumble in front of leadership, your impact goes down.

But when you’ve trained for it?

You’re calm. You pivot with ease. You leave the room knowing you earned trust, not lost it.

That’s exactly what our AI Training for MSLs and Medical Affairs teams (contact me) is designed to do. We teach your team how to turn unpredictable Q&A into your strongest moment.