The development team of Google Chrome he’s working on one new security feature based on the now ubiquitous artificial intelligence which, apparently, it could help you avoid online scams perpetrated by deceptive sites. Thanks to the use of an advanced linguistic model that acts locally, the Mountain View giant’s browser would be able to analyze the contents of the web pages visited by the user and identify any inconsistencies or danger signals. In addition, a review summary system for e-commerce it could help us determine the reliability of an online store by providing feedback based on reviews on platforms such as Trustpilot and ScamAdviser. Currently being tested on Chrome Canary (an experimental version of Google’s browser designed for developers and other professionals), it is not yet known when these innovations will see the light in the stable version of the browser.
How Chrome’s anti-scam system works and how to recognize fraudulent sites
In recent years, the number of counterfeit sites has increased exponentially, thanks to new technologies that make it increasingly easier to create convincing copies of authentic pages. These sites often imitate well-known brands or trusted services to trick users and steal personal or financial data (e.g., credit card number). To counter this phenomenon, Google is developing an innovative technology that exploits a LLM (Large Language Model) integrated into the browser, which will make it possible for the browser analyze the content of web pages directly on your device.
The system, identified by the leaker Leopeva64 and made known by the latter on X, is based on an experimental option called “Client Side Detection Brand and Intent for Scam Detection”. From the screens shared by the leaker, the functionality seems to be activated by placing a check in the Chrome Canary settings. Once the setting is enabled, the browser uses LLM to check whether the content of a page actually matches the stated intent and brand it represents. In other words, AI can detect discrepancies between what the site claims to be and its real intentions, thus helping the user avoid potential scams.
The function seems to be very careful in the respect user privacysince a LLM on-devicewhich means that the analysis of the web pages you visit takes place directly on your device, without sending data to the cloud to third parties.
What does the function to evaluate the reliability of online shopping sites consist of
In addition to the anti-scam system we just saw, Chrome could introduce a complementary functionality for online purchases. This is a dedicated button that, when positioned next to the address bar, will offer a AI-generated summary of store reviews. This data will be extrapolated from independent sources such as Trustpilot and ScamAdviser, enabling the user to have an immediate and quite reliable evaluation of the e-commerce site he is visiting.
