Google beefs up Gmail security to fight phishing attempts

The company is using machine learning to filter out spam and malicious content.

Google has just added a bunch of new security features in order to protect Gmail users from spam and phishing messages.

Though they didn't say as much, the bumped-up protection is likely in response to the phishing scam that went around earlier this month. The attack peddled a bogus Google Docs file in attempts to gain access to users' Gmail accounts. Google took measures to boost its security shortly after the incident, but the features announced today go even further.

First, Google has implemented a machine learning model to keep spam out of your inbox and to better detect phishing messages. The company estimates that between 50 and 70 percent of Gmail messages are spam and the new system blocks them with 99.9 percent accuracy. The machine learning system works with Google Safe Browsing to selectively delay emails (less than 0.05 percent of them) for added phishing analyses. It then generates URL click-time warnings for suspicious links.

Additionally, G Suite users trying to reply to someone outside of their company will get a warning asking if they really want to send the message. But users' existing and regular contacts will be picked up by Gmail so you're not bugged with excessive warnings.

Finally, Google also announced improved detection of malicious attachments in order to keep out sneaky ransomware and malware pushes, saying, maybe a little dramatically, "While the bad guys never rest, neither do we."