Online criminals have discovered new ways to scam people on the web. Google records an increase in fake voices, cloned faces and fake websites that copy real brands and lure victims into handing over cash or personal data.
UPI fraud alone drained INR 1,087 crore from Indian wallets in 2024, Google says. Forecasts place the 2025 loss at about ₹20 000 crore if defences stay the same.
To deal with the threat, Google has drawn up a Safety Charter that pairs machine learning with plain language advice. The charter treats every careless click, call or text as a weak spot worth fixing.
How Is Search Cleaning Up The Mess?
Google’s new Fighting Scams in Search report shows AI blocking hundreds of millions of risky pages each day. Better classifiers now catch twenty times more scam sites than before, the firm reports.
Criminals who pretend to run airline help desks no longer slip through so easily. Google records an 80% fall in that ploy after new filters went live.
All this checking happens in the blink of an eye, so genuine results still reach users with no delay.
Can A Browser Spot Danger Before You Do?
Chrome’s Enhanced Protection already cuts phishing and other traps in half compared with the basic setting, Google explains. The newest guard comes from Gemini Nano, a language model that runs on the device and reads each page as it loads. Because nothing leaves the machine, warnings pop up at once, even on brand new scam pages.
More from News
- What Are The Main Sources Google’s AI Overview Uses?
- New Drone Flights Approved to Help Monitor Railways
- How The UK Government Is Helping With Employment Reform
- What Are The Data-Related Risks Of Period Tracker Apps?
- Investment in UK Businesses Up 3% This Year
- How Much Water Does ChatGPT Actually Use?
- Why Is Tesla Facing Legal Action In Australia?
- How AI Is Helping Scammers Enrol Fake Students To Get College Funding
Gemini Nano now blocks remote tech support fraud, a common ruse where callers pretend a laptop is infected and then charge a fee to “fix” it. Google plans to carry the same shield over to Android.
The browser on Android also watches website notifications. When its on-device model spots a flood of shady alerts, it has the option for a single tap to unsubscribe or to view what was blocked.
On the phone side, AI in Google Messages intercepts more than 500 million suspect texts every month. A link from an unknown sender triggers an extra warning; last year, 2.5 billion such alerts flashed on screens.
The Google Phone app adds a live call monitor. If speech patterns match common fraud play books, the handset flashes a stop sign before any money changes hands.
Why Is India In The News?
Google’s DigiKavach campaign has already taught 177 million people how to spot online scams. The company now works with the Indian Cyber Crime Coordination Centre (I4C) to spread that reach some more.
Advertising scams also face more strict checks. Google removed 247 million misleading promos in 2024 and closed 2.9 million accounts that broke finance rules. Its verification rule means an ad for loans or shares appears only after the advertiser meets local law.
Google Play Protect scans more than 100 billion installed apps each day. In India, the pilot blocked nearly 60 million attempts to add high risk apps between October 2024 and May 2025, stopping more than 220,000 different titles on 13 million devices.
Users of Google Pay see pop ups when a transfer looks odd. Last year, 41 million alerts stopped payments that might have vanished into fraud rings.