Google today announced that the Google Safe Browsing service, which was first released for 2007, protects more than 3 billion devices. The number surpassed the 2 billions that the company said it protected in May of 2016.
Safe Browsing provides URLs containing malware or phishing content to Chrome, Firefox, and Safari browsers, as well as to ISPs. 
The service can also be accessed through one public API or directly by manually checking from this URL any site you want. OR Google has expanded its Safe Browsing to its various products over the years, such as Android, Ads, Analytics, Gmail, Google Play, and so on.
Keep in mind that, most of the time, Safe Browsing protects users without hindering them. This means you just do not see any malicious search results, Android apps on Google Play, questionable Gmail messages, etc., because they're just filtered. Other times, you're presenting an interstitial action like the one shown below (on mobile):
You can see the above image on pages that try to make you install malware or download some other unwanted software. They are automatically blocked in Chrome, Firefox and Safari, although you can view the site manually if you wish. Web developers and application developers (such as Snapchat) integrate Google Safe Browsing into their applications to check URLs before being presented to their users.
As for AI, Google Safe Browsing has been using mechanical learning for many years to detect various types of threats.
Google did not elaborate on how the project would evolve, but noted that the service had to tackle today's "web" and that the team was "constantly evaluating and integrating new approaches to improving Safe Browsing".
