Google AI Overviews are designed to give searchers a quick answer instead of making them work through a list of links. That convenience now has a sharper risk: scam phone numbers can appear inside AI-generated search summaries, where they may look authoritative enough to trust.
The problem is not just that AI answers can be wrong. In this case, a bad result can send someone directly to a person pretending to represent a real company, creating an opening for payment information or other sensitive details to be stolen.
How the scam reaches searchers
AI Overviews pull together information from across the web and package it into a synthesized response. That format can make the answer feel settled, even when the information behind it has not been properly checked.
Reports of scam support numbers appearing in Google AI Overviews have been spotted by The Washington Post and Digital Trends, with related examples appearing on Facebook and Reddit respectively. Credit unions and banks are also warning their customers about this type of fraud.
The basic path is simple. A person searches Google for a company name because they want a contact number. An AI-generated result shows a phone number. The person calls it, believing it belongs to the company.
But the call can instead reach someone impersonating that business. From there, the scammer may try to collect payment details or other private information under the cover of customer support.
Why AI summaries can make old scams feel new
Fake contact information has existed online for a long time. What changes with AI Overviews is how the information is presented. A number that might once have appeared on an obscure page can be lifted into a prominent answer box and displayed with the confidence of a direct response.
The source article says it is not clear exactly how the fake numbers are being inserted into the web. The best guess described there is that scammers publish them in multiple low-profile places online, paired with the names of major companies. AI Overviews may then gather that information and repeat it without enough verification.
That matters because many people treat a search summary differently from a regular search result. A link still asks the reader to judge the source. An AI answer can feel like the work has already been done.
This is where the danger sits. The interface can reduce the friction that normally pushes people to compare sources, check an official page, or notice that a result looks suspicious.
What Google says it is doing
Google says it is actively fighting scammers and continuing to improve spam-detection systems. In a statement to WIRED, the company said:
“Our anti-spam protections are highly effective at keeping scams out of AI Overviews and showing official customer support numbers where possible,”
Google also says it has recently launched updates intended to improve scam protections for AI Overviews specifically. Even so, the company encourages people to double-check phone numbers with additional searches.
That combination is important. Better filtering may reduce the problem, but users still need to treat contact details in AI answers as information to verify, not as the final word.
How to verify a number before you call
The safest rule is straightforward: do not rely on an AI Overview for phone numbers, payment instructions, or other specific details that could expose your money or identity.
If you need to contact a company, search for the company first and use the contact information listed on the company’s own website. That may take an extra click or two, but it gives you a better chance of reaching the real organization rather than someone copying its name.
Before calling, run another Google search for the number itself. This extra check can help you compare what the AI answer showed against other information on the web.
Be especially cautious if a customer service conversation turns toward payments or personal information. A real-looking phone number is not proof that the person on the line represents the company you meant to reach.
- Use the company’s own website for support numbers whenever possible.
- Search the phone number separately before calling.
- Treat AI Overviews as starting points, not confirmed records.
- Slow down around payment or personal data, even if the call seems routine.
When AI search is the wrong tool
There is currently no way to turn off AI Overviews. If Google displays one for a query, users can scroll past it or use a different search engine.
AI can still be useful for broad, conversational tasks. The source article gives the example of chatting with Gemini about ideas for a next vacation. That kind of exploratory use is different from looking up accurate information about hotels, cruise ships, or travel agents.
The practical line is this: use AI for ideas, summaries, and broad orientation, but be more careful when the answer contains a specific fact you plan to act on. Phone numbers deserve that extra scrutiny because one wrong call can put sensitive information in the wrong hands.
As search becomes more reliant on AI, the old habit of checking the source still matters. For contact details, the most useful answer is not always the fastest one. It is the one you can verify before you trust it.