For more than a decade, Google has used progress in AI to protect you from online fraud, where malicious actors are deceiving users to access money, personal information or both. Today, we release a new report on how to fight against fraud in the search, and share the new ways we use AI to keep you in safety across search, Chrome and Android.
Stop scam in search with AI-driven defense
In search, AI helps us discover and block hundreds of millions of scammy results every day. Our search fraud in Search Report shows how the investments we have made in our AI-driven scam detection systems together with improvements to our classifiers-have enabled us to capture 20 times the number of fraud pages. These improvements help ensure that the results you get are legitimate and protects you from harmful places trying to steal your sensitive data.
Progress in AI has strengthened our scam-struggle technologies-so we can analyze large amounts of text on the web, identify coordinated scam campaigns and discover new threats-offs one step ahead to keep you sure of the search. For example, we have observed a significant increase in bad actors on the web who mimic aviation providers and scams people in need of help. We have already reduced these scams by more than 80% in search, which greatly reduces the risk of calling a prisons phone number.
SuperCharging Safe Browsing in Chrome with Gemini Nano
The improved protection mode for safe browsing on Chrome is our browser’s highest level of protection and keeps users twice as secure against phishing and other fraud versus our standard protection mode.
We now use Gemini Nano, our on-Device Large Language Model (LLM) on desktop, to give improved protective users an additional defense against online fraud. The on-device approach provides immediate insight into risky sites and allows us to offer protection even against fraud that has not been seen before. Gemini Nanos LLM is perfect for this use due to its ability to distill the varied, complex nature of sites, which helps us to adapt to new scams faster.
We are already using this new AI-driven approach to protect users from remote technical support scams, one of the biggest online threats facing users today. Our goal is to expand this protection to Android devices and even more types of scams in the future.
Fighting scams, spam and unwanted messages
Sometimes the risk of scammy places can extend beyond the place itself. If you have enabled messages from sites, malicious sites may try to scam you through a blocking of notifications. To help you keep you in front of malicious, spammy or misleading messages, we launch new AI-driven warnings for Chrome on Android.
When Chrome’s Machine Learning Model flags a review, you will receive a warning with the possibility of either unsubscribing or viewing the content that was blocked. And if you decide that the warning was shown incorrectly, you can choose to allow future messages from this site.