The Planning Edge

A 2026 Study Tested AI Travel Planning.
Here's What It Found.

Forbes just published the data. The results confirm what every seasoned traveler should know before letting an algorithm plan their next trip.

Forbes · InsureMyTrip Study · March 23, 2026

A travel insurance company called InsureMyTrip recently tasked researchers with something straightforward: plan a luxury seven-day trip to Switzerland for four adults using three of the most widely used AI tools — ChatGPT, Gemini, and Google's AI Mode. The results, published in Forbes on March 23, 2026, are worth reading before you open another AI chat window to plan your next trip.

All three tools produced inaccurate plans. Not occasionally — consistently.

What the Study Found

The researchers evaluated each AI on accuracy, pricing realism, pacing, sourcing, and total travel time. The surface results looked impressive. The itineraries appeared polished, logically structured, and detailed. The problems emerged on closer inspection.

One restaurant recommended by ChatGPT did not exist. Michelin ratings were misrepresented across multiple tools. A flagship train experience — the Jungfraujoch journey — was listed at the Swiss franc equivalent of $129 to $258 per person. Actual rates start at $342. Hotel availability was not verified before inclusion. Restaurant opening hours were incorrect. None of the tools adequately accounted for the physical demands placed on older travelers across multi-day, high-elevation itineraries.

The researchers noted something telling: the confidence of the output did not reflect the accuracy of the information. The plans looked authoritative. They were not.

"The itineraries often looked polished and logical on the surface. Closer inspection revealed that both big and small details were often inaccurate."

Sara Boisvert, Director of Marketing — InsureMyTrip, via Forbes

Why This Happens

AI tools generate text based on patterns in training data. They do not call the restaurant to confirm hours. They do not contact the hotel to verify room availability. They do not access live pricing from suppliers. They produce a structurally coherent response — and the traveler discovers the gaps upon arrival.

This is not a criticism of AI as a technology. It is a description of what AI is and what it is not. It is a capable tool for generating frameworks and surface-level research. It is not a substitute for human verification, supplier relationships, or on-the-ground knowledge accumulated over years of working directly in the field.

The danger is not that the output looks rough. The danger is that it looks finished.

What Aurum Does Instead

Every venue, hotel, and experience Aurum includes in a plan has been confirmed — for existence, current operation, availability, accurate pricing, and fit for the specific traveler. Pricing comes from direct supplier relationships and live market intelligence, not training data from an indeterminate point in the past. Every day of an itinerary is calibrated to the travel party: their pace, their physical comfort level, their unstated preferences, and their priorities.

The Aurum planning fee exists precisely because this work takes time and expertise that cannot be automated. The alternative — a polished plan built on unverified information — costs nothing upfront and significantly more when things go wrong.

Switzerland received more U.S. visitors last year than at any point in the past three decades. If you are planning a trip there, or anywhere that requires precision, verified logistics, and a plan that holds up beyond the first read: that is what we do.

Ready to plan with confidence

Your trip deserves verified details.

Book a consultation. We'll discuss your travel goals and show you what planning at this level looks like.

Begin Your Consultation
Previous
Previous

Luxury Travel While Black: What You Need to Know

Next
Next

Why Your Room Category Matters More Than the Hotel