Fake Travel Itineraries Are Everywhere in 2026: How to Spot One Before You Book

On Halloween night in 2024, thousands of people packed O'Connell Street in Dublin waiting for a parade. It had a route (O'Connell Street to Christchurch Cathedral), a start time, and a convincing web page with photos and enthusiastic reviews. It lacked only one thing: existence. The listing came from a content-farm site called My Spirit Halloween that blended AI-generated text with photos lifted from real past events, it spread through TikTok, and in the end the Gardaí had to post a notice asking the crowd to go home. Nobody was robbed that night. They just spent an evening standing in the cold for an event no human had ever planned.

That is the fake-travel-content problem in miniature. In 2026, the itinerary you find on a blog, the guidebook you buy on Amazon, the villa you book from a search ad, and the "hidden gem" a chatbot recommends can all be produced by software with no connection to the physical world. Some of it is a deliberate scam. Most of it is just cheap filler published at scale. All of it fails the same way: it describes a world that does not quite exist, and you only find out when you are standing in it.

We have already written a five-minute routine for fact-checking your own AI itinerary before a trip. This post is about the step before that: how to tell whether the itinerary, guidebook, or booking site in front of you is grounded in reality at all, before you give it money or a morning of your holiday.

The four kinds of fake travel content circulating in 2026

AI-slop guidebooks. The New York Times documented this back in 2023: a "France Travel Guide" by one Mike Steves, a "renowned travel writer" with an AI-looking headshot, a biography suspiciously close to Rick Steves's, and around 100 five-star reviews, selling for $16.99 against the real Rick Steves book at $25.49. Mike Steves does not exist. The Times ran passages from dozens of similar self-published guides through an AI detector and the overwhelming majority came back as almost certainly machine-written, propped up by bought reviews. Three years on, the pattern has industrialised: search Amazon for any mid-sized city plus "2026 travel guide" and you will find shelves of them.

Invented destinations and attractions. Chatbots and content farms keep sending real people to places that are not there: travellers have driven for hours to soak in hot springs that were never built, and turned up to hike canyons that exist only in a model's training-data fog. The descriptions are vivid precisely because vivid text is what a language model is good at producing. Reality is not consulted.

Cloned booking sites. This is where the money is. McAfee's May 2026 research found that 38% of travellers surveyed had encountered a travel-related scam and 41% of those lost money, with Tripadvisor now the most impersonated travel brand, cloned at roughly three times the rate of Kayak, Expedia, or Booking.com. Fodor's, citing McAfee data, put losses from AI-powered travel scams at an estimated $13 billion, nearly $1,000 per victim. The wave has been building for a while: Booking.com's head of internet safety reported a 500 to 900 percent surge in travel phishing after generative AI tools arrived. The polished property page with lush photos and glowing reviews takes an afternoon to generate now, and it vanishes the moment your deposit clears.

Hallucinated itineraries for sale. Somewhere between the guidebooks and the scams sits a grey market of $9 itinerary PDFs, "local's guide" blog posts, and aesthetic TikTok day plans that were prompted, not lived. They are rarely fraud in the legal sense. They are just wrong: restaurants that closed in 2023, a "sunrise viewpoint" that turns out to be a car park, three neighbourhoods "ten minutes apart" that span the whole city.

How to spot a fake itinerary: a six-point checklist

The good news: fake travel content is easy to catch once you stop reading it as prose and start testing it as a set of claims. Ten minutes covers all six checks.

1. Put every named place on a real map

This is the single highest-value check. Open Google Maps and search for each specific place the itinerary names: the restaurant, the viewpoint, the "hidden courtyard café." A real place has a pin, photos uploaded by different people across different seasons, and reviews stretching back years. A fabricated place has nothing, or a suspiciously fresh listing with five photos in the same lighting. If an itinerary names eight places and two do not exist on a map, bin the whole thing: the author (human or machine) never checked, so nothing else in it deserves your trust.

2. Check opening hours and prices against the official source

Fake content loves confident specifics: "open daily 9 to 6, entry 12 euros." Pick the two or three anchor attractions in the plan and check them against the official website, not against another blog (which may have been generated from the same slop). The wrong closing day, a price that is years out of date, or a "skip-the-queue ticket" for a site that is free to enter are all fingerprints of generated text that no one verified.

3. Reverse-image-search the photos

Right-click any photo that looks a little too perfect and search with Google Lens. Three outcomes matter. The photo appears on the official site or across years of tourist uploads: fine. It appears on a stock-photo site or on pages about a completely different place: content farm. It appears nowhere at all, and the shadows are slightly wrong and the crowd's faces smear when you zoom in: AI-generated, and so, probably, is everything around it. The Dublin parade page worked because it used real photos of past events; a reverse search would have shown they came from a different festival entirely.

4. Do the logistics maths

Generated itineraries fail geometry constantly, because a language model knows place names, not distances. Look for the impossible commute: a plan that calls the walk from the Trevi Fountain to St. Peter's Basilica "a quick ten minutes" (it is about three kilometres and a river crossing, 40 minutes at a stroll), or dinner in one town and a sunset an hour's drive away, 20 minutes later. Plot the day's stops on a map and look at the shape. Real itineraries cluster; fake ones zigzag, because the model listed famous things rather than routing a human body through space.

5. Check the author's history

Before you buy a guidebook or trust a "local expert," spend two minutes on the byline. A real travel writer leaves a trail: older work, a site, interviews, social accounts with mess and inconsistency. Red flags: an author with dozens of destination guides published within months of each other, a headshot with waxy skin and mismatched earrings, reviews that arrived in one burst and all read alike, and a bio that mirrors a famous writer's just closely enough to draft on their trust. On Amazon, click the author's name and read the shelf, not the book.

6. Book only through channels you can verify

Never pay through a link that came to you: from an ad, an email, a chatbot answer, or a too-good listing. Type the airline, hotel or platform address yourself, or use the official app. Cross-check a property's name and phone number against its Google Maps listing. Pay by credit card, never by bank transfer (in the UK, Section 75 protection is exactly why), and treat urgency ("price locked for 14 more minutes") as the tell it usually is. A real deal survives an hour of checking; a cloned site depends on you not taking that hour.

Hold your AI planner to the same standard

Here is the uncomfortable part: a chatbot itinerary and an AI-slop guidebook are the same object. Both are fluent text generated from patterns, and both will invent a plausible restaurant with equal confidence. The difference between useful AI planning and slop is not the writing quality. It is grounding: whether every suggestion traces back to a record that exists in the world, with coordinates, opening hours, and years of human reviews attached.

So demand that standard from your tools. An AI travel planner is only trustworthy when every place it suggests is a real, verifiable map entity, not a name in a paragraph. That is the bet we made with Travolp: it plans from real Google Maps places, every stop lands as a pin on an actual map (so check number one happens before you ever see the plan), and the whole trip works offline once downloaded. No tool is immune to error, ours included, but a plan built from real records gives you a short list to verify instead of a wall of prose to investigate. Our comparison of AI travel planning apps looks at which tools ground their suggestions and which hand you a transcript, and if you are starting from scratch, our step-by-step guide to planning a trip with AI builds verification into the process from the first prompt.

The bottom line

Fake travel content is not going away; it is getting cheaper to make every month. But it all shares one weakness: it cannot survive contact with a map. Places that exist have pins, photos with history, opening hours on an official site, authors with pasts, and distances that add up. Before you buy the guidebook, book the villa, or walk out of the door with a downloaded day plan, run the six checks. The whole routine costs ten minutes. Standing on O'Connell Street waiting for a parade that was never real costs an evening, and the cloned hotel site costs a great deal more.