In terms of security, Chrome can now identify more than twice as many phishing attacks and dangerous websites according to Google’s most recent machine learning (ML) model, according to the company. The brand-new on-device machine learning model went live in March. According to Google, since it was released, Chrome has found 2.5 times more risks. Google has described how machine learning makes Chrome more enjoyable and safer.
Beyond security, Google may be preparing to employ machine learning to improve the experience of Chrome users.
Customers of Chrome can choose not to get alerts from pages they don’t want to see. Google claims it plans to integrate an AI model that learns from previous interactions when users are unlikely to approve prompts and may silence them to reduce distractions in the upcoming release of Chrome. When a website’s alerts are disabled, it will seem as follows:
Although the approach prevents interruptions for users, it does provide notifications if the ML model misreads the data (hey, it happens!)
The expansion of Google’s Journeys feature, which was introduced earlier this year, is the next item on the agenda. Journeys aim to help users use all the data that Google gathers about them to retrace their online steps. Google claims Chrome will compile all the pages you’ve visited for a specific topic by using some ML magic. The idea is to bring back the days of scrolling through our whole browser history to pick up where we left off in the past.
But the most intriguing part of the whole thing is undoubtedly the final step. According to Google, it will employ machine learning to instantly customize Chrome’s toolbar based on the user’s preferences.