Scammers have infiltrated Google's AI responses - how to spot them

Reddit / Elyse Betters Picaro / ZDNET

ZDNET's key takeaways

• Scammers are exploiting AI to trick people looking for customer numbers.
• Google's AI Overview, AI Mode, and OpenAI's ChatGPT are vulnerable.
• Run a regular search, or head to the company's website to find a number.

Get more in-depth ZDNET tech coverage: Add us as a preferred Google source on Chrome and Chromium browsers.

Do you ever use Google's AI-powered search to look for customer service numbers and other contact info? If so, you could be opening your wallet to a scammer.

Several people are revealing how they were scammed while searching for customer service numbers using Google's AI Overviews and AI Mode. Assuming the phone numbers that appeared in the AI-based results were legitimate, the victims called them and provided payment information to complete a transaction. Instead, the people behind the numbers were scammers trying to steal money from the callers.

Also: How web scraping actually works - and why AI changes everything

"I pride myself on being cautious," Alex Rivlin, owner and CEO of real estate firm Rivlin Group, said in a recent Facebook post, as described by The Washington Post. "I don't click links, I don't give personal info over the phone, and I always verify. But I still got caught in a very sophisticated scam -- and it all started with what looked like a legit phone number for Royal Caribbean I found on Google."

Rivlin revealed how the scammers were able to provide all the right information, including pricing, terminology, and shuttle details. He was then tricked into sharing his credit card information and realized he had been scammed after he discovered the bogus charges on his statement.

"I'm sharing this as a public service announcement," Rivlin added. "With AI-generated results and spoofed numbers, the game has changed. If it can happen to me, it can happen to anyone."

In another tale shared on Reddit, as spotted by DigitalTrends, a person had placed an order with Swiggy Instamart but hadn't received all the items. Using Google to search for "Swiggy customer care number," this individual called the number he found.

Also: Could AI help you finally escape the office? Most workers think so

The "customer care agent" on the other end tossed out legitimate questions. But then he asked the caller for his WhatsApp number, called that number, and told him to share his screen. That's when the customer got suspicious and ended the call. Swiggy itself doesn't offer phone support, only chat-based help.

How is this possible? 

Scammers have long been able to insert fake phone numbers and other info into search results, hoping to trick unsuspecting users into sharing account or financial details. But the problem has taken on a new twist due to the growing use of AI. With a traditional search, you'd normally find a host of results, most of which should be legit. But with AI, you're presented with a single overview or result, increasing the odds that you'll follow only that one.

The problem isn't limited to Google. Scammers have also exploited OpenAI's ChatGPT in much the same way.

(Disclosure: Ziff Davis, ZDNET's parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

To highlight one example cited by DigitalTrends, security experts at Odin and ITBrew recently explained how hackers can use prompt injection to add specific commands to Google Gemini. In this case, the command tells Gemini that it must include the scam message and the fake customer service number in its AI-based summary.

Also: How to get rid of AI Overviews in Google Search: 4 easy ways

Responding to a query from The Washington Post, a Google spokesperson said the company has "taken action" on several fake phone numbers and continues to work on improvements to address queries like these. OpenAI said many of the pages with fake numbers referenced by ChatGPT have been removed but that such updates can take time.

In response to ZDNET's own questions, Google shared the following statement: "We have strong protections and policies to prevent scams from appearing in AI Overviews or ranking highly on Search, and our systems are effective at surfacing official customer service information for the queries people search most. We've taken action on several of the examples shared, and continue to work on broader improvements to address rarer queries like these."

The best thing you can do

For now, the best thing you can do is avoid blindly calling any business phone number or contact information delivered by an AI. Use a more traditional search engine or mode where you can scour a variety of results. To do that, append "--AI" to your search query. Alternatively, head directly to the company's website to locate its number and contact details.

Further, here are some additional tips:

1. Check URLs before clicking on unfamiliar links. Hover over any link in an online ad to check the URL to make sure it's not suspicious or malicious.
2. Get information about the web page or online business. One way to learn more about a web page that pops up in a search is by using the "About this result" feature. Clicking the three dots next to a search result will tell you more about the source before you click its link.
3. Check for official resources. When you're looking for a customer service number or contact information, there's only one official online source: the company's website. Remember, not all businesses have customer service numbers. If you can't find one on the website, be wary of any numbers that appear in a search result.
4. Look for signs of a scam. When viewing any website, beware of odd formatting, unusual fonts, or unexpected characters. Those could be signs of a scam.
5. Highlight your customer support. If you run a business, highlight your customer support methods directly in Google Search. Doing so makes it easier for people to find the right way to contact you.