AI technology has rapidly become part of daily life, but it’s not flawless. It might seem like generative AI is omniscient, but it can make errors or even fabricate information, raising concerns among tech reporters like myself as companies integrate AI into widely used tools.
Search is a prime example. We’ve been trained since the late 90s to trust search results for information. Many of us use Google, enter a query, and accept the top results without question. Now, with Google’s AI Overviews dominating search results, users often accept AI-generated content at face value.
There are numerous issues with this new approach, with a major concern being that Google’s AI Overviews and AI Mode might suggest scam phone numbers when searching for a company’s contact information, as reported by Digital Trends.
Scammers manipulate Google’s AI. Digital Trends reported that Alex Rivlin shared on Facebook his experience with Royal Caribbean’s customer service. Unable to find their support number on their website, he googled it and called the number in the AI Overview. He unwittingly provided his credit card information to a scammer, recognizing the fraud only when asked for his birthdate and subsequently seeing suspicious charges on his card.
Currently, if you ask Google who owns that spam number, it incorrectly responds with Royal Caribbean, pulling data from a site impersonating a government page. Scammers appear to list fake numbers on bogus websites, tricking Google into using that data. AI sees “Royal Caribbean” next to a number on a .gov site, deems it authentic, and displays it in an AI-generated result.
Digital Trends shared a Reddit case where a user searched “how to fix a misspelled name on Southwest,” leading them to a scam number in an AI Overview result. The user noticed the scam after comparing it with Southwest’s legitimate number.
If you google these fake numbers, you’ll see results leading to misleading information designed to deceive Google into displaying scam numbers. Scammers target specific issues to increase the likelihood of their numbers appearing in relevant searches.
In another instance, reported by Times of India, a man lost over $3,400 after calling a fake number for Swiggy’s customer service, discovered through a Google search.
Never assume AI-provided answers are accurate. Google’s AI Overviews are flawed, struggling to discern legitimate from false information. It might include numbers from fraudulent sites impersonating official pages. This flaw was evident in Google’s problematic AI Overviews rollout last year, with the AI even taking jokes from Reddit as credible sources. Although AI can retrieve accurate information, its flaws make relying on AI answers risky. If you won’t check traditional links, at least verify AI source credibility. If a source seems dubious, so is the result.
For company contact details, always verify directly from the company’s website. If no phone number is listed, assume it’s unavailable and seek other official contact methods. Scammers can easily exploit open-web information.