Two kinds of AI, two very different jobs

Every article about AI in travel treats “AI travel tools” as one category. It isn’t.

There are two categories, and they answer two very different questions. Mixing them up is the reason so many travelers try one AI travel app, find it useless for what they actually need, and conclude that AI in travel is overhyped.

The two categories are strategic AI and tactical AI.

Strategic AI helps you figure out where to go and what your trip should look like. Tactical AI helps you win the transaction once you already know.

Both are legitimate. Both are useful. But they solve different problems, and if you pick the wrong one for the moment you are in, you will not get a useful answer.

Strategic AI: plan the trip

Strategic AI is what most people think of when they hear “AI travel planner.” It is the category Expedia just spent an undisclosed sum to buy into when it acquired Layla in July 2026 (Skift).

The strategic-AI category includes:

• Layla (now part of Expedia): identity-first trip planning, over 5 million users, 30 million conversations, and more than 2 million itineraries generated as of March 2026 (Layla, March 2026).
• Mindtrip: AI-first itinerary building with a media-rich planning canvas, expanded into booking in 2026.
• Wonderplan: conversational trip planning with a strong personalization layer.
• Kayak Ask AI and Priceline Penny: OTA-built conversational planners that sit on top of each brand’s inventory.
• Booking.com Smart Filter and Expedia’s new multi-agentic architecture: incumbent-scale AI planning built directly into the primary booking flow.

All of these tools do variations on the same job. They help a traveler describe a mood, a season, a budget, a set of preferences, and produce a trip idea, a shortlist, an itinerary. Some of them will also complete the booking. The differentiator across the category is the quality of the conversation and the size of the inventory the AI has access to.

Strategic AI is at its best when the traveler does not yet know where they want to go, does not know what kind of trip they want, or wants help thinking through options. It is the category that shows up when you type “ai travel planner” into a search engine, and it is the category that is currently absorbing most of the industry’s AI investment.

Tactical AI: win the transaction

Tactical AI is different. It doesn’t help you decide where to go. It helps you get a better deal, or a better seat, or a better answer, once the where and the when are already settled.

Tactical AI answers questions like:

• Is this flight a good price right now, or should I wait?
• Is there a better seat available on the plane I have already booked?
• Did the fare I paid drop after I booked, and can I get a refund?
• Which alerts should I trust when there are twelve of them in my inbox?

These are not conversation-friendly questions. They do not benefit from a chat interface that asks about your travel style. They benefit from a fast, confident, specific answer. “Yes, book it, this fare is in the bottom 10% of what this route has sold for in the last 90 days” is a tactical answer. “Have you considered visiting Porto instead?” is a strategic answer. Both can be useful. Only one is useful to a traveler who already has a boarding pass.

The tactical-AI category is much smaller than the strategic-AI category, and much less discussed. It includes:

• Hopper’s Price Prediction: the longest-running consumer example of tactical AI in travel. Predicts whether a fare is likely to rise or fall over the next week and gives a book-or-wait recommendation (Hopper Price Prediction).
• Kayak Price Forecast: similar prediction on top of Kayak’s historical pricing data, presented alongside search results (Kayak Price Forecast).
• Google Flights price tracking: alerts on flight-price changes for saved trips, with a lightweight “prices are typical / lower than usual / higher than usual” signal.
• ExpertFlyer: (business traveler-focused): seat-availability alerts and fare-class monitoring for travelers who care about upgrades and premium-cabin availability.
• SlickTrip: fare alerts for flights you have not yet booked and Better Seat alerts for flights you have already booked. Watches for price drops or preferred-seat availability and notifies you when something changes, leaving the booking or seat change in your hands. (SlickTrip)

That is roughly the extent of consumer-facing tactical AI in travel as of late 2026. Compared with the strategic-AI category, which now includes acquisitions worth hundreds of millions of dollars and every major OTA’s flagship product roadmap, the tactical-AI category is under-served.

Why the split matters

The reason the split matters is that the two categories are answering questions from two different moments in a traveler’s journey, and they cannot substitute for each other.

Ask a strategic AI whether now is the right time to book your Lisbon flight and it will offer to help you plan Lisbon. Ask a tactical AI where you should go this summer and it will not have an answer, because it is not that kind of tool.

There is a further consequence, and it is the one that matters for how travelers should think about which AI to use. Most of the questions travelers actually have during a trip are tactical, not strategic.

Consider the arc of a single trip:

1. Where should I go? Strategic. Asked once, at the very beginning.
2. When should I go, and can I afford it? Bridging. Asked once or twice.
3. Is this a good fare? Should I book now or wait? Tactical. Asked repeatedly, often across days or weeks.
4. Did the price drop? Is there a better seat? Did the schedule change? Tactical. Asked continuously from booking until departure.
5. Where should I eat, what should I see, what should I skip? Bridging back to strategic, in-destination.

Most AI-in-travel coverage focuses on the first question. Most of a traveler’s actual AI use, if the category matures, will be spent on the third and fourth.

How to tell them apart

A quick test. If you can answer the question you are asking without knowing where you are going, you need tactical AI. If you cannot answer it without knowing where you are going, you need strategic AI.

• “Is $487 a good price for LAX to LIS on October 12?” → tactical.
• “Where should I go for a beach week in October, budget $2,500?” → strategic.
• “Is there a middle-of-the-plane window seat open on UA 907?” → tactical.
• “What are the best three days to spend in Lisbon?” → strategic.

The consumer AI travel category is well-covered on the strategic side. On the tactical side, it is mostly still open. That is the category worth watching.

What comes next

Both categories will grow, but they will grow differently. Strategic AI will consolidate. The Layla acquisition is the first of what will likely be several. Expect Booking Holdings, Priceline, and Tripadvisor to make similar moves in 2026 and 2027, and expect the surviving strategic-AI products to be the ones with the largest inventory access and the deepest personalization signal.

Tactical AI will proliferate. The barrier to entry is lower, the addressable question set is broader, and no single incumbent has locked up the category. SlickTrip is one of the products being built for the tactical column, with a forthcoming ChatGPT app for seat alerts, fare alerts, and travel bucket list already in OpenAI’s app review process (SlickTrip for ChatGPT). Expect more.

The travelers who get the most out of AI over the next three years will be the ones who understand which category answers which question, and who use both.