Google Messages trains artificial intelligence to detect spam using signals from unencrypted chats

Google Messages introduces a new way to fight spam and uses data from unencrypted chats to train artificial intelligence.

Here's What We Know

The updated algorithms now better recognise spam by analysing URLs and other signals in messages, which helps to avoid unwanted and harmful messages.

Google Messages' spam protection feature is enabled by default and automatically filters suspicious SMS messages to the spam folder. The data from these messages is now used to improve AI models, which helps to identify new types of spam. Encrypted messages, such as RCS chats, remain untouched and are processed only on the device.

The update also allows you to flag spam not only for Google, but also for the mobile operator. This helps to better filter fraudulent messages. It is not yet known how these changes will affect users.

Source: Google