“ChatGPT Ruined My Trip”: The Nightmare of Tourists Who Blindly Trust AI

No more dog-eared old guidebooks and dozens of tabs open on the computer. Nowadays, one in four French people—and even one-third of those under 35, according to the Kantar Institute—is considering entrusting the organization of their vacation to artificial intelligence. The promise is admittedly enticing: all you need to do is provide a budget and a few preferences, and the tool will spit out a detailed itinerary in seconds.

After continually delegating the logistics of their holidays to algorithms, many tourists end up disappointed.AI is a travel guide that is sometimes extremely arrogant and capable of lying with great confidence.

The nightmare of false recommendations

The major problem with these virtual assistants has a well-known technical name: hallucination. When AI does not know the answer, it prefers to invent one rather than admit its ignorance, and in the field of tourism, the consequences are disastrous.

In a recent study conducted by the British agency SEO Travel, one hundred travel itineraries were generated by ChatGPT. The result is damning. Almost all of them contained at least one error, and in a quarter of cases, the bot enthusiastically recommended restaurants or attractions that had definitively closed for good. This is exactly the misadventure that Sophie, 31, experienced in Italy.

After asking for a family-run spot away from the tourists, she set off enthusiastically. “I was even pleased because I thought I had found a gem that the tourists did not know about,” she told BFMTV. But after a twenty-minute walk, cold shower: there was nothing but an ordinary residential building. Yet the AI had assured her that the establishment was “particularly renowned for its fresh pasta.”Furious, the young woman realizes the full extent of the deception: “She had made everything up”.

Administrative errors that cost a lot

If going in circles in search of a nonexistent restaurant is irritating, AI hallucinations can have dramatic financial and administrative consequences. Spanish influencer Mery Caldass learned this the hard way. Filmed in tears in her boyfriend’s arms, she recounts her fiasco: “I asked ChatGPT whether I needed a visa; it said no, and I didn’t check. I will never trust this twerp again.” The tool had indeed “forgotten” to tell her that an electronic travel authorization (ESTA) was strictly required to travel to Puerto Rico.

Content creator Chaimae Lebyed is similarly disillusioned. Thinking she could get her visa on the spot in Fiji, as the AI had firmly assured her, she ended up stuck at the airport on the day of departure. A disastrous failure that cost her 700 euros. “I’m supposed to be at the airport, and I can’t go because ChatGPT has ruined my trip”, she fumes in a video, before bitterly exclaiming: “What’s the name of the OpenAI CEO? Give me my money back”.

When AI Puts Hikers in Mortal Danger

It is probably out in the great outdoors that artificial intelligence proves most dangerous. For more than a year, national park professionals have seen tourists arriving with completely unrealistic AI-generated itineraries. “It started last summer. We are seeing more and more people coming in to have an AI itinerary validated,” warns Adrien Majourel, communications manager for the Cévennes National Park.

At this same park, staff have seen Parisian hikers arrive with a ChatGPT-generated schedule planning 82.5 km of walking in two days—a route that normally takes four—with elevation changes underestimated and recommendations to bivouac in strictly prohibited areas. Given the increasing number of these risky behaviors, the park even had to publish an unequivocal warning: “Artificial intelligence does not walk on trails. Preparing a hike is not simply drawing a line on a map.It’s about adapting an itinerary to the weather, to a fitness level (...), and for that, nothing replaces the human touch. »

While artificial intelligence can prove to be an excellent tool for finding inspiration or doing initial groundwork, it must never have the final say. Between fake rental listings generated by scammers and chatbot hallucinations, human verification and seeking out local professionals remain, more than ever, the true guarantors of successful holidays.

👉🏻 Follow the latest tech news in real time: add 01net to your sources on Google and subscribe to our WhatsApp channel.