The Invisible Revolution: Why AI Is Rewriting the Rules of European Tourism

Reflections from EU Tourism Day 2026 on how artificial intelligence is transforming not just technology, but traveller behaviour itself, and what Europe must do to lead rather than follow.

Five years ago, travellers searched.

Today, travellers ask.

Tomorrow, travellers will delegate.

This simple observation was the starting point of my remarks at EU Tourism Day 2026 in Brussels, during the panel « Making the European Tourism Offer Meet Demand. » It sounds almost banal. But the implications are profound, and largely misunderstood.

We tend to frame the AI conversation around technology: large language models, generative interfaces, recommendation engines. But the real disruption isn’t happening inside the algorithms. It’s happening inside traveller behaviour.

The biggest change is not the tool. It’s the expectation.

For two decades, travellers have been trained to search. To type keywords, scroll through results, compare options, read reviews, and make decisions. The entire tourism industry – destinations, OTAs, hotels, DMOs – has been optimised for this behaviour. Websites, brochures, SEO strategies, advertising campaigns: all designed to capture attention in a world where travellers actively browse.

That world is ending.

Today, travellers increasingly ask rather than search. They prompt AI assistants with natural language: « Where should I go for a quiet weekend in Southern Europe in late October? » The interface is no longer a list of links. It’s a conversation. The traveller doesn’t want ten options. They want one good answer.

And tomorrow? Travellers will delegate entirely. Agentic AI systems will book trips, manage itineraries, and make decisions on behalf of users based on inferred preferences, past behaviour, and contextual constraints. The traveller won’t even see the options that were filtered out.

This is the quiet rupture. Demand is no longer explicit. It is inferred.

According to Phocuswright’s recent research, 39% of travellers now use AI for researching and planning trips. Among Millennials, that figure is 58%. Among Gen Z, 45%.

And according to KAYAK’s WTF report, 41% of Gen Z and Millennial travellers already trust AI recommendations more than advice from people or social media.

Read that again. A plurality of young travellers trust the algorithm more than their friends.

We have entered what I call the « no-click era. » Travellers don’t want to click through to your website. They want the AI to tell them whether your destination is worth visiting, without ever visiting its website.

AI is becoming the first interface between demand and offer.

Not the website. Not the booking engine. Not the travel agent. The AI.

This changes everything. Destinations and businesses are no longer competing only for the traveller’s attention. They are competing for the AI’s understanding.

If your destination’s data is poorly structured, incomplete, or outdated, the AI will either ignore you or misrepresent you. If your offer isn’t semantically clear, if it can’t be parsed by a machine reasoning about context, you become invisible.

Visibility is no longer about marketing spend. It’s about machine-readability.

Europe has the richest tourism offer in the world. History, culture, landscapes, gastronomy, diversity, no other region comes close. But this richness is also our vulnerability.

European tourism is fragmented, local, and heterogeneous. Thousands of small destinations, family-run businesses, and regional specialities. This is what makes Europe unique. But it’s also what makes Europe difficult to structure digitally.

Most of the European tourism offer is still designed for a world of search, not a world of AI mediation. Websites, campaigns, and even destination strategies are optimised for human browsing, not machine reasoning.

The risk is clear: if we don’t structure our offer ourselves, global AI systems will do it for us (i.e. from USA or China…). Using partial data. Biased signals. Commercial shortcuts. Optimising for volume and short-term conversion, not sustainability or local value.

The opportunity is equally clear: Europe can become the first region where tourism supply is not only digitised, but intelligible, interoperable, and policy-aligned.

Meeting demand no longer means attracting more tourists. It means matching the right demand with the right place, at the right time, under the right conditions.

This is the shift from volume to precision. From growth at all costs to sustainable matching.

AI makes this technically possible. We can now process behavioural signals at scale, understand intent in real time, and dynamically adjust offers based on context. But this only works if destinations move:

• From static offers to dynamic ones. Your destination isn’t the same in August and November. Your offer shouldn’t be either.
• From marketing metrics to behavioral signals. Clicks and impressions tell you what happened. Intent signals tell you what’s about to happen.
• From dashboards to decision systems. Data for reporting is no longer enough. We need data that drives action—automatically, in real time.

The question every destination should be asking: Are our contents ready for Agentic AI?

And this brings us to the heart of the matter.

Who trains the AI? Who feeds it? Who governs it?

Because AI is only as good as the data that feeds it. And right now, most tourism AI is fed by the same sources: major platforms, dominant OTAs, anglophone reviews, and commercially incentivised content.

This creates bias. It creates blind spots. It creates a system where some places are over-recommended, and others become invisible, not because of quality, but because of data availability.

Artificial Intelligence is becoming a strategic infrastructure for the tourism industry. It shapes visibility, access, and flows at scale. It determines which destinations appear in AI-generated itineraries, which businesses get recommended, and which experiences are surfaced.

That makes it a form of power.

If this power is used only to optimise volume and short-term demand, the result is predictable: pressure on residents, saturation of iconic places, and the gradual loss of social license for tourism itself.

We’ve already seen this pattern. Venice. Barcelona. Dubrovnik. Amsterdam. The backlash against overtourism isn’t about tourism per se; it’s about tourism that ignores local limits, resident needs, and long-term sustainability.

AI could make this worse. Much worse. At scale.

Or it could make it better.

In Europe, we can choose a different path. We can use AI to actively balance interests: visitors and residents, growth and livability, competitiveness and sustainability.

That means embedding policy objectives directly into how demand is matched with supply, by design.

Not as an afterthought. Not as a filter applied post-hoc. But as a core logic of the matching system itself.

This is what I mean by « from promotion to orchestration. » Destinations don’t just market themselves. They actively manage flows, distribute demand across time and space, and use AI as a tool for stewardship rather than just acquisition.

This requires action at multiple levels.

• For destinations: Invest in structured, machine-readable data. Build content architectures that AI systems can parse. Think about your offer not just in terms of what humans see, but what machines understand.
• For businesses: Move beyond traditional digital marketing. Understand how AI recommender systems work. Ensure your data is complete, accurate, and semantically rich.
• For policymakers: Recognise AI as strategic infrastructure. Support interoperability standards. Fund data commons. Ensure that European tourism data serves European policy objectives—sustainability, territorial cohesion, and resident wellbeing.
• For all of us: Start the conversation about AI governance in tourism. Who should control the training data? What values should be embedded in the matching logic? How do we ensure transparency and accountability?

These are not technical questions. They are political questions. And they require political answers.

In the age of AI, competitiveness is not about being visible everywhere. It is about being correctly understood.

The destinations that will thrive are not those with the biggest marketing budgets. They are those whose offer is clearly structured, whose value proposition is semantically precise, and whose data is rich enough for AI systems to represent them accurately.

This is the new battleground. Not attention, but comprehension. Not impressions, but inference.

Europe has the content. We have the stories. We have the experience. What we need is the architecture – digital, institutional, and political – to make that content intelligible to the machines that are increasingly making decisions on travellers’ behalf.

The invisible revolution is already underway. The question is whether Europe will shape it, or be shaped by it.

I know which future I’m working toward.