Google says deep learning helped Gmail detect malicious documents

Spam e-mails are a menace and Google is using deep learning models to help identify and block such content from reaching your inbox.

Gmail's existing machine learning models help block more than 99.9 per cent of threats from reaching inboxes. The company recently added a new generation of document scanners that relies on deep learning to improve detection capabilities, the official blog notes.

The improved scanner, launched at the end of 2019, increased daily detection coverage of Office documents containing malicious scripts by 10 per cent, and improved detection rate by 150 per cent. It uses TensorFlow deep-learning model trained with TFX and a custom document analyser for each file type.