AI Trip Planning: What ChatGPT Can (and Can't) Do For You
AI has completely changed the way people plan travel — and honestly, I love using it. It's fast, it's fun, and it's full of ideas. But there's something every traveler should know before they start typing prompts into ChatGPT: it's an incredible brainstorming tool. It is not a travel agent.
What Happened With One Client's Australia Trip
I recently had a client use AI while planning a trip to Australia. To be fair, it came up with some genuinely fantastic ideas — places to visit, experiences to consider, and smart ways to structure the itinerary.
But there were also some big problems.
One suggestion was a Tauck itinerary that sounded perfect on paper — except it was completely sold out for the entire year. The estimated flight costs were also off by a wide margin, coming in at roughly a third of the actual airfare.
It also suggested ending the trip with three nights in Kakadu. A lovely idea in theory — but my trusted local partners flagged that the timing fell relatively early in the Top End season, when access can depend heavily on how the preceding wet season played out. Roads and visitor sites aren't always guaranteed to be open. As one example, Ubirr — one of Kakadu's most iconic sites — remained inaccessible over Easter in 2025 due to high water levels. AI has no way of knowing that. It's simply not designed to factor in seasonal, on-the-ground conditions like this, which is exactly why we'd want to plan around them rather than assume everything will be accessible.
That gap matters more than people realize. When the real pricing and availability finally come back, it can look like the travel professional is the one overcharging or getting it wrong — when in reality, the AI simply never had access to live inventory or current fares in the first place.
The truth is simple: AI can generate information. A travel professional validates it.
It's Not Just Australia: More Places AI Gets Trip Planning Wrong
The Australia trip isn't a one-off. These are the kinds of gaps I see regularly when clients bring me an AI-built itinerary, across nearly every type of travel:
- Small-ship and expedition cruises. AI will happily recommend a specific cabin category on an Antarctica, Galápagos, or river cruise sailing — but it has no idea that the popular categories on the best departures often sell out 12 to 18 months in advance. It also can't tell the difference between a "guarantee" category (you're assigned a cabin later) and a specific, chosen cabin, which can matter a lot for accessibility, noise, or motion sensitivity.
- Safari lodges and camps. AI might suggest a beautiful camp without knowing it closes entirely during the rainy season, or that the experience itself changes dramatically by season — some camps switch between fly-in access and long road transfers, and river levels can make boat safaris impossible for weeks at a time. On paper it looks bookable. In reality, it may not even be open.
- Hotel room categories and resort fees. AI tends to compare hotels by a single nightly rate pulled from a search result, without factoring in resort fees, local taxes, or whether breakfast and transfers are included. It also can't tell you that the specific "ocean view" or "suite" category it recommended is already sold out, even if the hotel itself still shows as "available" online.
- Visa and entry requirements. Entry rules change more often than people expect — a country can shift from visa-on-arrival to a mandatory e-visa, add a reciprocity fee, or change vaccination requirements with little notice. AI's information can lag behind these changes, which is not something you want to discover at check-in for your flight.
- Promotion stacking. AI will sometimes suggest combining two "current" promotions it found in its training data or a search result, without knowing that suppliers rarely allow the same booking to combine both offers — or that the promotion expired months ago.
Where ChatGPT Genuinely Shines
Used the right way, AI is a wonderful starting point for a trip. Here's what to lean on it for:
- Destination inspiration — "What should I see in Australia?"
- Comparing destinations side by side
- Building a wish list of experiences you're excited about
- Understanding the differences between regions, hotel brands, or cruise lines
- Figuring out the right questions to bring to your travel advisor
Where ChatGPT Falls Short
Where AI runs into trouble is anything tied to real-time, real-world logistics — because it simply doesn't have access to it. Here's what that looks like in practice:
- Real-time hotel or cruise availability. AI's training data has a cutoff, and even web-connected tools aren't plugged into live supplier inventory systems the way a travel advisor is. A sailing or hotel category can look wide open in a chat response and be fully booked in reality.
- Current airfare pricing. Airfare changes by the hour based on demand, fuel surcharges, and fare class availability. AI-generated estimates are often based on stale averages, which is exactly how a client can end up expecting a fare that's a third of the actual cost.
- Whether a specific room category is actually available. A hotel showing availability on its website doesn't mean every category is open — the specific suite, view, or bed configuration you want may already be gone, something only a live availability check can confirm.
- Current promotions or combinability. Suppliers run promotions on their own calendars, with their own fine print on what can and can't be combined. AI has no visibility into whether an offer is still active or whether it can be paired with another rate.
- Visa and entry requirements without verification. Entry rules, reciprocity fees, and health requirements change with little notice and vary by nationality and length of stay. AI can give you a helpful starting point, but it should never be your final word before you book flights.
- Exact flight schedules. Airlines adjust routes, timing, and equipment regularly, and schedules more than a few months out are especially prone to change. An itinerary built around a flight time that later shifts can throw off an entire trip's connections.
- Seasonal access and site conditions. Weather-dependent road closures, wet-season impacts, and park or site accessibility (like Kakadu's Ubirr) depend on real-time, on-the-ground conditions that no AI model has access to.
- What is actually bookable today. At the end of the day, an itinerary is only useful if it can be turned into confirmed reservations. AI can describe a wonderful-sounding trip that simply isn't purchasable in its current form.
This Is Where Your Travel Advisor Comes In
This is exactly the gap a luxury travel advisor is built to close. I have access to booking systems, supplier relationships, current inventory, contracted rates, promotions, and air departments — plus the experience to recognize when something that looks perfect on paper isn't actually the best option for you.
But access to systems is only part of the story. What really makes the difference is what I bring to the table before a single reservation is ever made.
A black book built over years, not minutes. I've spent years curating a network of trusted partners around the globe — people on the ground in the destinations I sell, who have their finger on the pulse of their region at all times. They know before anyone else when a new lodge opens, when a hotel undergoes a renovation, when a road washes out, or when a camp's best season shifts. That kind of real-time, local intelligence simply isn't something AI can replicate.
Hotels that are constantly, personally vetted. My partners are always evaluating properties, both new and established, to make sure they're truly right for my clients — not just well-reviewed online, but right for the way my clients actually like to travel. A hotel can look stunning in photos and still miss the mark on service, room size, noise, or location. My partners catch that before you ever arrive.
Restaurant lists built from firsthand tastings, not star ratings. The restaurant recommendations I hand my clients aren't pulled from a review aggregator — they come from partners who have actually eaten the food, at that table, recently. That's the difference between "highly rated online" and "I sat there last month and it was extraordinary."
Scouting trips that put me in the room. Our team regularly travels specifically to research destinations, properties, and experiences firsthand, so our recommendations come from real knowledge rather than a brochure or a search result. We've slept in the sheets, dined in the restaurants, and experienced the service level ourselves— so when we tell a client a property is worth it, it's because we've lived it, not because an algorithm scored it well.
A network of advisors who share what they know. Beyond my own experience, I'm part of a network of fellow travel advisors who collaborate and share ideas constantly. If I haven't personally been somewhere, chances are someone in my network has — or they've sent clients there and can pass along detailed, honest feedback on how it actually went. That kind of collective knowledge is worth its weight in gold, and it's built from years of relationships in this business, not something you can prompt your way into overnight.
That's the difference between a dream and a deliverable.
My Favorite Way to Use AI
Let AI help you dream. Then let your travel professional make the dream real.
You don't have to choose between AI and your travel advisor — use them together. Bring your ChatGPT wish list to me, and I'll turn it into a real, bookable, beautifully planned trip with pricing and availability you can actually trust.
Ready to turn your AI-inspired trip into reality? Contact Jen Mitchell Travel to start planning.