Fake Hotel Listings Are Getting Harder to Spot: Can an AI Detector Help Travelers?

Imagine arriving in London after a long flight. Your bags are heavy, everyone wants dinner, and your hotel address leads to a pub.

This is no longer a far-fetched travel nightmare.

In September 2026, staff at The Liberty Bounds pub near Tower Bridge were turning away as many as 20 travelers each day. Scammers had used the pub’s address for a fake apartment listing on Booking.com. Travelers arrived expecting accommodation and found a Wetherspoons instead.

Another fake listing used Portcullis House near Parliament as the address for apartments close to Big Ben. Some travelers reportedly reached the government building before learning no apartment existed there.

Those stories came weeks after Which? Travel deliberately listed 10 Downing Street as a vacation rental. The experiment was designed to test Booking.com’s verification system.

The listing got through. Fourteen people asked about the property during a 20-minute booking window. A Which? researcher also completed a test payment for the fake rental.

For travelers, there is an uncomfortable lesson here. A professional booking page does not automatically mean the hotel behind it exists.

Fake Listings No Longer Need to Look Fake

Older travel scams sometimes gave themselves away quickly. Photographs were poor, sentences were strange, and payment requests arrived from suspicious addresses.

Scammers now have far better tools available. Generative technology can produce believable property photographs and polished descriptions within minutes. An AI detector may help you identify some synthetic images or writing before you book. Still, it cannot answer every important question.

NIST says the accuracy of synthetic-image detection varies according to the detector and the material being examined. Compression and the image generator used can also affect detection results. A detector trained around one generator may perform less successfully against another system.

This means a detector result should start your investigation. It should never finish it. There is another problem travelers should understand.

Many scammers do not need AI-generated photographs at all. They can copy genuine pictures from a real hotel and attach them to a false address or fake booking page.

Your detector might correctly identify those photographs as real. The hotel listing could still be completely fraudulent.

The 10 Downing Street Test Was Surprisingly Easy

The Which? investigation is worth understanding because the property chosen was hardly obscure. Researchers submitted 10 Downing Street as a one-bedroom apartment in central London. They included its actual address and a photograph of the famous front door.

Booking.com did not immediately stop the listing. Which? reported that proof of identity or ownership was not required before the property initially went online. Under the platform rules examined by Which?, some identity checks could come months later.

Researchers even posted a fake 10/10 review afterward. Booking.com told Which? that its test did not represent the experience across millions of properties. The company also said its security measures remove the majority of fraudulent listings within 24 hours.

Still, travelers do not need millions of fake listings. One fake property can ruin one carefully planned vacation.

Rental Scams Are Already Costing Travelers Millions

The United States has its own numbers showing the scale of fraudulent rental listings.

The Federal Trade Commission reported nearly 65,000 rental-scam complaints between 2020 and June 2025. Reported losses reached about $65 million during that period. The median reported loss was $1,000.

Younger travelers were particularly exposed. People aged 18 through 29 were three times more likely than other adults to report losing money through a rental scam.

Many fake listings begin with something attractive. You may find a better location than expected. The price may be noticeably below nearby hotels. Beautiful photographs can make the deal harder to ignore. Then the pressure begins.

A supposed owner may request money quickly because another traveler is interested. Someone may send a payment link through a message. Others may ask for a bank transfer outside the booking platform. Those requests should slow your booking process immediately.

How an AI Detector Can Actually Help

An AI detector works best as one check among several. Suppose a villa photograph has strangely shaped windows or impossible reflections. Running the picture through a detection tool might give you another reason to investigate. You can then check the image elsewhere.

Google Lens allows users to search for similar photographs and websites containing related images. This can help you discover that your supposed hotel photograph actually belongs to another property.

For a suspicious listing, try these checks before paying:

• Search the complete street address separately.
• Find the hotel through its own official website.
• Call the property using an independently found number.
• Compare photographs across several established booking platforms.
• Use Google Lens on important property photographs.
• Check the location using an online map.
• Read recent reviews across more than one website.
• Keep every payment inside the official booking platform.

These checks take several minutes. They can save an entire trip.

Payment Requests Deserve Extra Attention

Scammers like payment methods that give travelers little chance of recovering their money.

The FTC advises travelers to avoid sellers demanding payment through wire transfers, gift cards, cryptocurrency, or certain payment apps. Those methods can make recovering lost funds extremely difficult.

The FTC also recommends typing a company’s known website directly into your browser rather than trusting a sponsored search result. Scammers can purchase advertisements that resemble legitimate hotel or airline results.

Unexpected links deserve the same caution. A message inside a booking platform can still contain a dangerous external link. Which? demonstrated this during its Downing Street experiment by sending an outside payment link through Booking.com’s own messaging system.

A message arriving through a familiar platform is therefore not enough proof.

Check the Address Before You Pack

This might be the simplest travel habit worth adopting. Put the hotel’s full address into a map before departure.

Check what businesses are actually registered there. Compare exterior photographs with those shown on your booking page. Search the property name separately from the booking platform.

If the listing claims to be a luxury apartment but the address belongs to Parliament or a pub, you have your answer.

Travelers sometimes skip this step because a booking confirmation gives them confidence. The recent London cases show why that confidence needs checking.

AI Detection Helps, But Verification Helps More

There is a temptation to search for one tool that can tell us what is genuine. Travel scams are unfortunately more complicated.

An AI detector can help identify material that deserves another inspection. NIST itself continues testing how reliably automated systems can separate AI-generated content from human material. Its evaluations show that some generators can still fool detection systems. Your best defense is therefore a collection of small checks.

Inspect the photographs. Search the address. Call the property. Compare prices elsewhere. Keep payments on recognized platforms.

Most importantly, give yourself time before sending money. A beautiful hotel page can be built quickly. Your vacation takes much longer to plan. Spending ten extra minutes checking where you will actually sleep is time well spent.

Want to discover more hidden gems and helpful travel tips? Join our free newsletter for the latest travel secrets and travel articles.

We are reader-supported and may earn a commission on purchases made through links in this article.