Over the last few months, something interesting has started to happen at KMS Travel.
Alongside the usual channels, like our B2B partners, Google and other browsers, referrals, private travel advisors, trade fairs and our own commercial efforts, we are beginning to see travelers and new partners mention that they found us, or saw us recommended, through tools such as ChatGPT, Gemini and other AI-assisted searches.
For a local DMC in Chile like us, that is not a minor detail. It means AI is changing how people plan travel, but more importantly, how they discover whom to trust. That may sound like a small distinction, but I think it matters. Much of the conversation around AI and travel still moves between two extremes: either AI will replace everyone, or human travel advisors will somehow remain untouched. I do not find either view very convincing.
A recent The Wall Street Journal article, republished in Chile by El Mercurio, caught my attention precisely because it pointed in a more interesting direction: travel advisors are not disappearing. In some segments, especially more expensive and complicated trips, they seem to be gaining relevance again, and from what we are seeing, that makes sense.
I am not trying to predict the future of the whole travel industry. I can only speak from the segment KMS Travel Chile knows best: tailor-made, high-value journeys to Chile and South America, especially for long-haul travelers from markets such as the United States. And in that segment, AI is not making trusted local expertise irrelevant. It is making the difference between generic intermediation and real expertise much more visible.
The Savile Row comparison
The clothing comparison helps me think about this: excellent prêt-à-porter exists. It is efficient, elegant and right for many situations, but it did not eliminate the tailor for some segments. The tailor still has a place because some clients want something made with them in mind. Someone takes measurements, listens, adjusts, corrects and remains responsible for the result, and travel has a similar logic.
AI will improve (and has already improved) "the prêt-à-porter side of travel". It will make standard planning faster, more accessible and probably better than many current booking flows. That is good. But some journeys still need the equivalent of a tailor: not because the traveler cannot find information, but because they want "warm" interpretation, care and responsibility with a face. For a high-value trip, especially to a destination far from home, the human conversation is not a luxury add-on. It is often part of the reassurance that allows the traveler to move forward or, at least, to "break the ice" in the initial parts of the commercial process.
Somewhere between hype and maturity
We are still in the "Middle Ages" of AI in travel: there is a bit of everything: useful tools, exaggerated promises, real productivity gains, weak recommendations, impressive demos, hallucinated details, better search experiences and plenty of noise. In that sense, the comparison with the early dot-com years is tempting. There was a real technological shift, but also a lot of confusion about what would actually last. Something similar is happening now. Some AI tools will disappear. Others will quietly become part of normal workflows. And many of today’s weaknesses, like poor context, outdated information, factual errors, biases, generic answers, etc., will gradually be better understood and reduced.
For travel companies, the practical question is not whether to “believe” in AI or resist it. The question is much simpler: where does it really help, and where does human judgment still do a better job? I see AI as very useful in the early and repetitive parts of the process. It can answer basic questions, help travelers explore possibilities, draft initial replies, sort information, and reduce the time spent on routine work. Used properly, it should give humans more time for the part of the job that cannot be reduced to a fast answer.
That is not a defensive view. It is actually a fairly optimistic one, because a good itinerary is not only a list of places. Travel planning has always involved information. But high-value tailor-made travel is rarely just an information "problem". A traveler can ask AI for the best time to visit Patagonia, how many days to spend in the Atacama Desert, or what to include in a first trip to Chile. In many cases, the answer will be useful. It may even be a good starting point.
But once a real trip starts to take shape (or is in progress and "something unexpected occurred"), the questions become different:
- Does this route make sense in that season?
- Is this hotel really the right fit for these travelers?
- Is this transfer too tight?
- Is the experience worth the extra night?
- What happens if the weather changes the plan?
- Which supplier is reliable not only on paper, but in practice?
- Where is it better to slow down rather than add one more highlight?
Those questions are harder (yet) to answer warmly from a distance. Chile, in particular, rewards good planning. It is a long and logistically diverse country, with very different travel realities between Patagonia, the Atacama Desert, the Lakes District, Chiloé, Rapa Nui, the wine valleys, the fjords, the Carretera Austral and Antarctica-connected journeys. A map can make it look simple. Anyone who has operated trips here knows it is not.
This is where the local DMC still has a role that is difficult to automate. Not because AI is useless, but because the final value often lies in judgement, timing, supplier knowledge and the ability to say, honestly, “I would do this differently.”
The visible local counterpart
Furthermore, there is a human layer that is nowadays easy to underestimate until a client actually needs it. In our experience, many travelers still appreciate a call. Some directly ask for one. Travel companies and advisors want to know who will be receiving their clients in the destination. Direct clients want to understand who is behind the proposal, especially when the journey involves a serious investment of time and money.
For us, the United States is our main market and often the one associated with the highest-ticket journeys. These are not usually simple purchases. They are long-haul trips, often planned months in advance, with expectations around comfort, safety, pacing, authenticity and support. At that level, a proposal is not only a document. It is also a promise or a commitment by someone. That is why the local DMC with a visible face still matters: the person or team who knows the suppliers, has worked with the guides, understands the practical limits of a route and can answer when something changes. This is not about romanticizing the "old-fashioned" travel model. Some parts of that model were slow, opaque or simply inefficient. AI will improve many of those weaknesses, and it should. But there is a difference between making the process more efficient and removing the human responsibility behind a complex trip.
AI is also a discovery channel
One of the more interesting changes is that AI is not only helping travelers plan. It is also starting to influence who gets found. For years, travel discovery depended heavily on Google rankings, paid campaigns, OTAs, marketplaces, referrals, trade shows and brand reputation. All of that still matters. But AI-assisted search is adding another layer, for when a traveler asks ChatGPT, Gemini or Claude for a reliable local specialist in Chile, the answer is not built only from ads or traditional rankings. It seems to draw from a wider set of signals: content, clarity, reputation, consistency, authority, reviews, structured information and how well a company is understood online.
That is why being recommended by AI tools is encouraging, but also demanding. It suggests that local specialists with real knowledge can become more visible in this new environment. It also means weak or generic positioning will be easier to expose. If everyone can produce surface-level content faster, then surface-level content becomes less useful. The companies that will benefit are probably those that have something real to say, and enough operational substance behind it. We are aiming for that.
The pressure will not be the same for everyone
This is where I think the conversation needs more nuance: AI will almost certainly put pressure on parts of the travel industry that depend mainly on controlling search, traffic, comparison or simple transactions. Steve Endacott has written about how AI search could challenge Tripadvisor and OTAs by changing how travelers discover, compare and eventually book travel. Whether the full scenario unfolds exactly that way or not, the direction is worth taking seriously. If a traveler can ask an AI assistant to compare hotels, flights, reviews, weather, prices and availability in one place, the value of some traditional intermediary layers becomes less obvious.
For "simple, on-a-budget trips", this may work very well. An "affordable" hotel stay, a standard city break or a basic flight-and-hotel combination may become easier to plan with very little human involvement, but that is not the whole travel market. A tailor-made journey through Chile or South America is often a different kind of decision. It involves geography, timing, local operations, personal taste, risk management, expectations and trust. It also involves emotional weight: anniversaries, family trips, once-in-a-lifetime journeys, retirement travel, honeymoons, adventure travel or a long-planned visit to Patagonia, Atacama or Rapa Nui. At that point, the question is no longer only “can this be booked?” but “who do I trust to make this work well?”
A more useful way to think about AI
For local DMCs like us and specialist agencies, the future should not be framed as human versus AI. That feels too simplistic. The better question is how to become more useful with AI than without it. AI can help us answer faster, reduce repetitive work, manage first contacts more smoothly, improve internal knowledge, support marketing, identify patterns and appear in new discovery channels. But it should also push us to a "happy problem": sharpen what is truly ours: destination knowledge, supplier relationships, honest advice, operational experience and a human counterpart who can be reached. In that sense, AI is not lowering the standard for local specialists, but raising it. The companies that only add a layer of generic intermediation will probably feel more pressure. The companies that combine technology with real local expertise may find themselves in a stronger position than before.
That is what we are beginning to see from our own corner of the industry. AI is already changing how travelers search. The next question is who they will trust once they find an answer.
