Google Chrome will incorporate an AI-powered scam detector for Android phones.

Google has been implementing various privacy and security features in its Chrome browser for a considerable amount of time. A notable example is the Enhanced Safe Browsing feature, launched in 2020, which provides protection against harmful websites and files by detecting threats in real-time. Three years after its introduction, Google has changed this feature from optional to a default security protocol, aimed at protecting users from phishing attacks, harmful extensions, and malicious downloads.

Recently, the company has started deploying its AI, Gemini Nano, to safeguard smartphone users from potential online fraud, particularly those that disguise themselves as technological security warnings on websites. This tool is designed to detect patterns of tactics used by fraudulent sites that display false alert messages. When a website exhibits risky behavior, Gemini Nano identifies the true intent behind the warning and communicates it to the Safe Browsing system. If a site is flagged as potentially dangerous, the browser will send a notification to users.

The notification will include the website's name along with a message warning about its suspicious nature. In addition to the option to unsubscribe with a single tap, users also have the choice to add it to a whitelist in case Chrome's new security feature fails. This feature will first be available on Android smartphones by the end of this year, though Google has not specified the hardware or software requirements for its implementation. According to the company, this feature is a priority on Android because most notifications are sent to mobile devices; however, its expansion to other platforms will also be evaluated.

The importance of this initiative lies in the rise of tech support scams, where sites present alerts warning that the user's device is infected with a virus. Google has observed that these fraudulent sites often present themselves in various forms, making it essential to assess the threat as users perceive it. Until now, Google relied on a system that paired blocked content in real-time and shared a small sample of browsing data for security reasons. However, with the arrival of Gemini Nano, security scanning is done on-device, which represents a crucial advantage. The company has pointed out that, due to the average lifespan of a malicious site being less than 10 minutes, on-device protection allows for the detection and blocking of attacks that have not yet been tracked.