Circle to Search on Android can assess whether a suspicious text message is a scam. You circle the message on screen and Google shows an AI Overview (the AI-generated summary that appears at the top of search) with guidance and suggested next steps. Google highlighted this use on its official blog, in the safety and security section, in October 2026.
For the person receiving the message, it is a quick check before tapping a link. For anyone running a brand that scammers like to imitate, it is a warning: the answer about your name now comes from an AI summary built from what is published.
How do you use Circle to Search to check a message?
Circle to Search checks a message in three steps, according to Google:
- With the message open, long-press the home button or navigation bar on your Android phone to activate Circle to Search.
- Draw a circle around the message you are unsure about.
- Read the AI Overview that appears, with guidance and next steps.
If the result shows red flags, Google recommends deleting the message and blocking the number. The point is to check before tapping any link or sharing personal information, which is exactly the moment a scam works.
Is Circle to Search always right?
No. Circle to Search returns an AI Overview, and that kind of summary can be wrong, like any AI answer. Google itself presents the result as “guidance”, not a verdict.
The bigger risk is the false negative. A new scam, with wording that has not circulated online yet, is less likely to be recognized. A neutral answer does not mean the message is safe. The old rule still applies: banks, carriers and stores do not ask for passwords, verification codes or urgent payments by text.
I worked at Banco do Brasil and saw the pattern up close: the scam that works is not the most sophisticated one, it is the one that arrives when the person is in a hurry. A ten-second check breaks exactly that hurry, and that is where the feature earns its keep.
Why does this matter for brands scammers imitate?
It matters because Circle to Search’s assessment is an AI Overview, and that summary leans on what the web says about the subject. If your company is imitated in text scams (delivery, billing, banking, telecom, e-commerce), the answer your customer sees depends partly on whether clear, official information exists about how you communicate.
A brand that never published “we do not send payment links by text” leaves the answer to forums, complaints and third-party news. It is the same logic behind how AI search gives old negative news new life: the summary uses what it finds, not what you wish it found.
The practical part fits on one page of your site:
| What to publish | Why it helps |
|---|---|
| Official contact channels, with real numbers and sender names | Gives the AI a source to compare the message’s sender against |
| What the company never asks for by message | Becomes a clear sentence that is easy to cite in a summary |
| Examples of recent scams that used the brand | Helps people recognize the text going around |
| How to report it and who to contact | Covers the next step the customer needs |
Write it plainly, with question-style headings (“Does the company send payment links by text?”). That is the format AI answers tend to pick up, the same one I cover in the guide on how to show up in ChatGPT and other AI assistants. And if your brand already has negative mentions tied to scams, the playbook for recovering from negative brand mentions applies too.
Who can use the feature?
The feature works on Android phones that support Circle to Search. Google’s post describes how to use it but does not list countries, languages or device models in the main text. If long-pressing the navigation bar opens Circle to Search on your phone, you can try it now.
Summary: Circle to Search against text message scams
- Circle to Search on Android assesses suspicious text messages: circle the message and read the AI Overview with guidance.
- If red flags appear, Google recommends deleting the message and blocking the number.
- The answer is an AI summary and can be wrong; a neutral result does not guarantee the message is safe.
- For brands imitated in scams, the answer depends on what is published. An official page with real channels and what the company never asks for by message gives the AI a source to use.
What to do now: try the feature on a spam message you have already received, so you know how it responds. If you look after a brand, search Google for “scam” plus your company name and see who is answering for you today. If it is not your own site, the page in the table above is your next job.