Booking.com warns of up to 900% increase in travel scams

Booking.com is warning artificial intelligence (AI) is driving an explosion in travel scams.

The firm's internet safety boss, Marnie Wilking, said there had been "anywhere from a 500 to a 900% increase" in the past 18 months.

She said there had been a particularly marked increase in phishing - where people are tricked into handing over their financial details - since generative AI tools like ChatGPT burst onto the market.

"Of course, we've had phishing since the dawn of email, but the uptick started shortly after ChatGPT got launched," she said.

"The attackers are definitely using AI to launch attacks that mimic emails far better than anything that they've done to date," she said.

Phishing attacks often try to convince people to hand over their card details through by sending them fake - but very convincing looking - internet booking links.

Scammers often target websites like Booking.com and Airbnb because they allow people to list their own places to stay

After someone pays up, the scammers either vanish without a trace - leaving the buyer without a place to stay - or even try to scam them out of more money through follow-up messages.

These sorts of scams have been around for decades, though they often come with telltale signs of fraud, such as spelling mistakes and grammatical errors.

But - speaking at the Collision technology conference in Toronto - Ms Wilking said AI was making them harder to detect because it could generate realistic images and much more accurate text, in multiple languages.

She is calling for hotels and travellers to use two-factor authentication - it involves an additional security check, such as inputting a code sent to you phone - calling it "the best way to combat phishing and credential stealing".

She also urged people to more vigilant than before when clicking on links.

But despite criticising how scammers are using AI, she said the technology was also allowing Booking.com to rapidly remove fake hotels that tried to scam people.

"We've set up AI models to detect those and either block them from getting on there to begin or take it down before there's any booking," she said.