Google uses AI to make email smarter

Google Inc. is applying its artificial intelligence (AI) technology to not only prevent spam emails but also as an email reply tool for users, a company engineer said Wednesday. Google said its machine learning model can ferret out spam and phishing messages with 99.9 percent accuracy, even spam that are carefully crafted to deceive people, "Our goal is to make Gmail the most productive email in the world," said Paul Lambert, Google's product manager. Defining five main threats to Gmail as malware, account hijacking, phishing, web attacks, spam, the engineer said Google aims to keep users safe. The new system uses an AI-based neural network to analyze and flag the suspicious messages, especially those written specifically to subvert simpler filters. Google estimates up to 70 percent of messages in Gmail's inbox are spam, Lambert said.