What just happened? Google does a good job of fighting junk emails. The company stops more than 99.9 percent of spam, phishing, and malware from ending up in Gmail users’ inboxes through the use of AI and other protections. Now, it’s improving the process even further, thanks to its TensorFlow machine learning framework, which is blocking an extra 100 million spam messages every day.
TensorFlow, which Google made open source back in 2015, has been used to improve spam detection for a while, but it's now received broader adoption and a larger set of spam classifications “at scale,” said Neil Kumaran, product manager at Google's Counter Abuse Technology division, speaking to VentureBeat.
Google is blocking the extra 100 million spam emails by identifying categories that used to be hard to detect. TensorFlow is used to stop image-based messages, emails with hidden embedded content, and messages from newly created domains.
100 million emails might sound like a lot, but when put into context against Gmail’s 1.5 billion users, it only works out at one extra blocked spam email per 10 users, according to The Verge. It’s still an impressive achievement, of course, especially when you consider how many unwanted messages it already blocks.
Google says that TensorFlow’s machine learning capabilities also help it personalize spam protections for each user. What one person thinks of as junk, such as newsletter subscriptions, might be considered important to someone else.
“There’s no one definition of spam out there,” added Kumaran. “But AI could help work out the best definition for you.”