AI Is Killing the Boring Parts of Being a Travel Agent

Rita Carton, a Los Angeles-based travel advisor, ran nine simultaneous trips in Q1 of last year. The year before, she could manage three. Nothing changed about her hours. What changed was that AI now handles the draft, the research, and half the email queue — and she handles the client.

Meanwhile, 44% of travel agents in a 2025 global survey of more than 1,300 professionals said they believed AI could eventually replace them. Among those with fewer than three years in the field, the number climbed to 51%.

These two realities exist at the same time. That's not a paradox — it's a fork. The difference between the advisors who are growing and the ones who are anxious isn't experience, or specialty, or even which tools they've bought. It's a decision that happens before the client call: which parts of this job only I can do, and which parts should I hand off?

But knowing two outcomes exist doesn't tell you which pressures created them. That answer starts with the consumer data — and it arrived faster than most advisors noticed.

The Floor Is Dropping Under One Part of Your Job

The threat to travel agents isn't a single AI platform announcing it can replace them. It's quieter and more specific than that.

Fifty-six percent of active U.S. travelers used AI for trip planning, booking, or in-destination help in the prior twelve months — up from 33% in early 2025. Among those users, 94% used it specifically for travel. A separate survey found that 37% of American travelers are now turning to AI for destination recommendations, itinerary creation, restaurant suggestions, budget planning, and general research. Those are tasks that have historically been the primary reasons a client picked up the phone.

This is where the pressure actually lives. Not in some future where AI books entire trips autonomously, but in the present, where a significant portion of your potential clients are doing their own first-draft research before they ever contact you.

David Shull, CEO of Tern — a travel-tech firm with usage data from 6,900 advisors — puts the internal math bluntly. He estimates that 80 to 90% of an advisor's day is administrative tasks that don't add unique value or use their expertise. "It's not actually adding unique value to their client," he said. "It's not using their expertise."

Most travel advisors don't see the need, or even get offended, when AI is brought up in the context of travel planning. But I couldn't disagree more. It's not replacing us.
— Athena Livadas, Owner, Páme Travel

If 80 to 90% of your day is work AI can do, the question isn't whether that time is at risk. The question is what you do with it when AI takes it.

This applies regardless of specialty. The tasks under pressure — research, comparison, first-draft itineraries, standard email responses — are the same whether you book luxury safaris or group cruises. The question is what remains once those are handed off.

The Three-Panel Map of Your Workday

The advisors growing right now are not avoiding AI. They're using it to compress the 80% of their day that doesn't require them, so they can spend more time on the 20% that only they can do.

What AI is already doing for you: Rita Carton estimates AI cut her average trip-planning time by 50%, halved her client response time, and freed the capacity that let her handle nine simultaneous trips in a single quarter instead of three. She uses AI for draft research, proposal generation, and email scaffolding — then applies her own judgment before anything reaches a client. The time AI saves is only valuable if you redirect it toward work only you can do.

What AI still can't do: Trish Smith, a Kansas City-based group-travel advisor, keeps a specific example ready when clients arrive with ChatGPT itineraries: the actual transfer time from Osaka station to the hotel while dragging luggage, because there are no taxis or rideshares at that exit. A language model will get the itinerary right and the logistics wrong. Smith uses AI for research, emails, and images — but she describes it to clients as her "office assistant," and she corrects it. She calls the expectation that someone can "plug something into AI and get an itinerary in 50 seconds" and immediately have it booked the "microwave society." Her value isn't the draft. It's knowing what the draft got wrong.

We do not use A.I. to replace any of our skills but use it to replace tedious work such as organizing notes, creating itineraries, creating itinerary outlines.
— William Lee, Travel Advisor and Head of Marketing, Chima Travel

If you know things AI cannot know — transfer realities, seasonal nuances, accessibility requirements, supplier relationships — those facts become more valuable precisely because AI is making generic itineraries abundant.

What becomes more valuable because of AI: Marie Russell, founder of Confetti Travel Service in North Carolina, recently spent two months coordinating flights and hotels for nine travelers from three different departure cities, all flying together to Asia and returning separately. "AI couldn't have even touched what I put together for them," she said. The value wasn't information. It was the coordination of competing constraints under ambiguity. No model has a client's real budget ceiling, their sister-in-law's mobility needs, and their tolerance for overnight layovers loaded together.

The more AI can produce a plausible plan, the more valuable a human becomes for cases where plausible isn't good enough.

The advisors who are shrinking are treating their job as information delivery. The ones growing are treating it as judgment delivery, with AI handling the information layer.

The Honest Reckoning

Here's where the data gets uncomfortable — because not every advisor is positioned to make this transition, and the labor trends show exactly where the floor is already dropping.

A Stanford study using ADP payroll data covering millions of U.S. workers found no evidence of widespread economy-wide job displacement. But employment of workers aged 22 to 25 in AI-exposed occupations was 19% below where it would have been had it kept pace with less-exposed peers. The decline operated mainly through reduced hiring, not separations. In travel, that translates directly: junior research and coordination roles face a thinning pipeline. Experienced advisors working complex, relationship-heavy accounts show no comparable gap.

The trust gap matters too. Despite high adoption numbers, a Booking.com report found that 91% of respondents still had concerns about AI and only 35% fully trusted its outputs. A YouGov survey of 14,000 adults across 12 European countries found that 59% were not yet ready for fully AI-driven booking. That gap is a structural opening for human accountability — but only for advisors who make their accountability visible.

"It's the travel advisor with AI that replaces the travel advisor without AI," said Arman Bimatov, co-founder of TravelWits. This is not a reassurance. It's a competitive threshold.

The Bureau of Labor Statistics projects travel-agent employment to grow modestly through 2034 — slower than average, but growth, not collapse. The real redistribution is happening inside the occupation: toward specialists, toward advisors with relationships and judgment, toward those with AI infrastructure. The floor is dropping for generic information providers. The ceiling is rising for everyone else.

If your client base is primarily people who want a hotel booked and a flight confirmed, AI is already a direct competitor. If your client base includes people who need someone to hold the plan together when things go wrong, that work is structurally resistant to automation — and it is becoming more valuable.

Three Moves, Starting This Week

Rita Carton didn't triple her capacity by finding a new niche or rebranding. She tripled it by deciding, at a specific moment, which parts of her job only she could do — and letting AI handle the rest. Here's what that decision looks like in practice.

This week: Run a task audit on your last five client interactions. List every task you completed and ask one question about each: could a well-prompted AI have produced a usable first draft of this? Any task where the answer is yes is a candidate for compression. Start with the three most repetitive — draft inquiry responses, destination research summaries, and proposal reformatting. The goal this week isn't to find the perfect tool. It's to see where your time is going that doesn't actually require you. This takes 20 minutes.

This month: Build one workflow that AI cannot replicate. Pick a trip type you handle regularly and create a structured discovery conversation — a set of questions that surface what the client actually wants versus what they think they want. Trish Smith asks clients considering Jamaica whether they're a beach bum, a party animal, or an adventure seeker. One set of answers sends them to Seven Mile Beach in Negril. Another sends them jumping into a waterfall in Ocho Rios. No AI itinerary generator starts there. Write your version of that conversation down, use it with every client for 30 days, and refine it. This is your verification layer — the thing that makes your recommendation different from a ChatGPT draft.

This quarter: Make your expertise findable by AI. Travelers increasingly start trip planning in a conversation with an AI assistant rather than a search engine. If your website, social profiles, and online content don't describe your specific expertise in plain, direct language — "In eight years of booking multigenerational Italy trips, I've found that..." — AI tools cannot recommend you when a potential client asks who to trust with that exact kind of trip. Audit your digital presence for one specialty. Write one piece of content that describes what you know, not just what you sell.

The advisors who are growing right now are not different from the ones who are struggling in experience or expertise. They are different in timing. They decided, before the pressure became obvious, that AI was infrastructure — like the phone, like the GDS, like the internet — and that their job was to decide what to build on top of it.

The question was never whether AI would change travel. It already has. The question is whether your clients find the new version of you — or find someone else first.

Recommended Tools & Resources