Google's AI overviews showing scam support numbers

There have been multiple cases of fraudsters using AI to scam consumers in new ways. With hidden phishing messages to voice impersonations, the recent incident cautions people looking for support numbers and calling numbers they are served via AI overviews. For a lot of Google searchers, AI overviews are the first thing that pops up whenever they search for anything.

Shared by Alex Rivlin on Facebook (via Digital Trends) who was trying to book a shuttle for his Royal Caribbean cruise to take him from the city center to the port. Since Royal Caribbean didn’t have an online booking system, Rivlin ended up searching for their customer service number on Google. The first thing that popped up was the AI overview showing a summary for his query, along with a support number.

Upon calling the phone number, Rivlin was greeted by someone representing Royal Caribbean and they provided all the right information, along with asking for the credit card details to make the payment for the shuttle. Then, the representative asked about paying a dock fee to which Rivlin agreed, but that was followed by a few personal information questions. He declined to give any more information than needed and after getting off the call, noticed some unsanctioned charges on the credit card. Fortunately, he was able to speak with the credit card company and block those charges without losing any money.

This wasn’t the first such scam. A few months ago, a Reddit user shared a similar experience where the support number for Southwest Airlines shown by Google’s AI overview turned out to be a scam number.

With fraudsters finding more sophisticated ways of using AI to scam consumers, it is becoming increasingly challenging to trust AI overviews and AI search results. In some cases, even cross referencing the support number shows up as a legitimate company customer service number. Furthermore, for a lot of users, the AI overviews are the first to show up when searching Google. The key takeaway is to be extra cautious about information served by AI overviews and careful about sharing any personal information with said numbers.