6 Ways AI Can Ruin Your Vacation (and How to Avoid It)

Some travelers use AI to help them budget, but find that AI-powered search engines often struggle with the details. “Many people don’t realize, but AI often hallucinates when doing simple addition,” says Daniel Herszberg, the co-founder of Travel Insighter, who has visited every country in the world. Herszberg also finds that generative AI may be great at adding up big-ticket items like hotels, flights, and food. However, it often overlooks other costs, including airport transfers or local trains between attractions. Other times, Herszberg finds that AI simply invents numbers that may appear correct at first glance. Mistakes in simple addition or made-up costs can completely derail your budget.

Fortunately, Herszberg has caught most budgeting errors before finalizing his trips, but others haven’t been so lucky. “I could just as easily imagine being stuck in a Rome train station and realizing that I’m a few hundred Euro short,” he says.

The fix: While AI can help create a beginning budget, always dig into the details and triple-check your numbers.

AI-generated or edited photos are often gorgeous but wind up misleading travelers. “We’ve seen travelers arrive at destinations with expectations shaped by AI-generated images where the skies are bluer, the crowds are gone, and everything looks picture-perfect,” Barnes says. When they arrive and “reality doesn’t match that fantasy, it can lead to real disappointment,” he explains.

Sometimes, travelers have the opposite problem and assume a photo of a gorgeous location is AI-generated when it’s not. Suspicious travelers may not book a trip as a result. The problem is pervasive enough that Icelandair launched an ad campaign letting travelers know that they don’t use any AI images in their advertising, which showcases the beauty of Iceland. “But when you’re looking at images of spectacular landscapes, breathtaking skies, and the bluest of waters, you might find yourself asking–is this real? In Iceland’s case, the answer is yes,” Icelandair states on its website.

The fix: Cross-reference any too-good-to-be-true travel photos with images of the real thing. You should be able to quickly determine whether a photo is real or created, wholly or in part, by AI.

AI tools that let travelers edit photos quickly and easily are making it harder to get a sense of what a destination is really like. There is a “weird trend now where AI is ‘fixing’ travel photos to make them seem calm and pristine,” even if a place is actually chaotic and packed with tourists, Herszberg says. Until recently, convincingly removing objects from photos took a long time. Now, “Phones have an AI ability to erase objects from the background of photos, and it is so quick and easy to do,” Barnes says.

That means it’s simple for nearly anyone to remove garbage bins, construction sites, and other tourists from a photo in seconds, he explains. As a result, even if you see a photo posted by someone you know who has recently been on a trip, you still can’t trust that what you see is an accurate representation. “It’s not the end of the world, but it does set you up for disappointment when the real thing doesn’t match the fantasy version the internet sold you,” Herszberg says.

The fix: Don’t choose a vacation destination or set expectations based on a single photo. Look at several images and think critically about whether what you know about the destination matches what you see. If someone posts a photo of themselves in front of the Colosseum at the height of peak tourist season and no one else is there, it has probably been tweaked. That doesn’t mean you shouldn’t go, but it does mean you won’t be disappointed when you show up and find a crowd.

Using AI to get visa advice may seem like a helpful shortcut. However, it’s not always accurate. “Visa requirements are a classic example of AI getting it wrong,” Herszberg says. For example, “You’ll see AI confidently say ‘China is visa-free,’ which is technically true for some nationalities or for certain transit routes. Yet, if you’re American, you still need a visa,” he says. AI may not tell you that, Herszberg says.

Additionally, AI may also not give you the whole story, especially if you don’t plug in the right query. Relying on AI for visa advice recently ruined a Spanish influencer’s trip. The influencer used ChatGPT to research whether she needed a visa to visit Puerto Rico, a US territory. ChatGPT correctly told the influencer that she didn’t need a visa. However, the AI search engine did not inform the influencer that she needed another type of travel document, an Electronic System for Travel Authorization (ESTA). Because she relied on AI and didn’t ask the right question, the influencer had to cancel her trip.

Alternatively, AI may tell you that you can get a visa on arrival when you actually need to apply for a visa in advance, Durflinger explains. Conversely, due to administrative requirements, it sometimes makes more sense to get your visa on arrival, even if AI recommends getting one prior to arrival, Barnes advises.

Visa mishaps can put an end to your trip before it begins, so it’s essential to verify any visa-related information well in advance of travel.

The fix: Always consult the official website of the country you are visiting for accurate, up–to–date visa information.

It’s tempting to plug your destination and the length of your trip into an AI-powered search engine and wait for the perfect itinerary to appear. However, that rarely happens. “People will use AI to give a sample itinerary, for example, build me an 8-day budget-friendly trip to Egypt, including all of the highlights,” Barens says. However, AI often suggests activities that are too jam-packed and don’t allow enough time at each location or a lunch break, he says.

Plus, AI often doesn’t account for real-world conditions, such as planning indoor activities during the height of the afternoon heat, time spent in long queues to enter an attraction, or rush-hour traffic, he explains. Another problem is that AI might recommend visiting sites in an illogical order, such as visiting one museum in the morning, going to lunch an hour away across the city, and then traveling back to hike in a park next to the museum in the afternoon.

The problem is pervasive. Andrew van den Broeck, private guide and brand manager at Yellow Zebra Safaris, finds that he is receiving more and more requests from customers who relied on AI for advice before booking. “It’s evident that our on-the-ground expertise and firsthand knowledge are often being weighed against the information clients gather through AI-driven searches,” he said. van den Broeck is used to clients conducting extensive research before contacting Yellow Zebra. This research can be valuable and help Yellow Zebra staff create a traveler’s ideal safari. However, since AI burst onto the scene, he has observed that the itineraries clients are requesting are increasingly difficult, or impossible, to book for a variety of reasons that on-the-ground agents can spot immediately. “This shows us that our consultants are still integral to the research, planning, and booking process, and AI is not replacing the human connection element of planning a safari trip,” van den Broeck says.

The fix: AI can be a good starting point for planning a trip. However, be sure to tweak your schedule so you aren’t running yourself ragged and have time to enjoy the places you’re visiting. Speaking to a travel agent with local expertise can also help you sort through an AI-suggested itinerary to create a realistic schedule.

Even if an AI-generated itinerary initially seems well-paced and reasonable, it may still be problematic. “We’ve had guests mention restaurants, viewpoints, or even ‘hidden spots’ that AI tools recommended only to discover they’re closed, inaccessible, or don’t exist at all,” Barnes says. Moreover, AI often uses outdated or incomplete information. That means “It can miss critical details like seasonal closures, weather disruptions, or cultural considerations that impact a trip,” such as changes in opening hours or festivals related to a local holiday, he adds.

Relying on outdated or inaccurate information isn’t only inconvenient. It can be dangerous. “In one instance, two travelers in Peru followed AI-generated directions to a ‘hidden’ Incan landmark that didn’t actually exist. They spent hours trekking through the mountains before realizing the destination was a complete fabrication,” Durflinger says. In that instance, a local guide overheard the travelers discussing the non-existent site and stopped them before they headed into a potentially dangerous situation. Had a local not stopped the travelers, they could have gotten lost in a remote, inaccessible area or experienced health issues related to the high altitude. Not everyone chasing a site hallucinated by AI will be as lucky if they aren’t careful.

The fix: Always double-check information provided by AI. Ensure a place you are interested in exists and will be open when you want to visit.