Google has stepped in to make it clear that the newly introduced Android System Safetycore app does not perform client-side scans of content.
“Android provides many device protections that protect users from threats such as malware, messaging spam and abuse protection, phone fraud, and more, while maintaining user privacy and keeping users in control of their data.” When I contacted.
“Safetycore is a new Google Systems service for Android 9+ devices that helps users detect unwanted content. Through optional and enabled features.”
Safetycore (package name “com.google.android.safetycore”) is part of a set of security measures designed to combat scams and other content that are considered sensitive to Google messaging apps for Android, and is now available in October 2024. It was first introduced by Google on the month.
Features that require 2GB of RAM are deployed on all Android devices, running Android version 9 or later, and running Android Go, are lightweight versions of the operating system for entry-level smartphones.
Client-side scanning (CSS), on the other hand, is seen as an alternative approach to enabling device analysis of data, rather than weakening encryption or adding background to existing systems. However, this method raises serious privacy concerns as the time for abuse is ripe for abuse by forcing service providers to search for material beyond the scope originally agreed upon.
In some ways, Google’s sensitive content warnings for the messaging app are very similar to Apple’s communication safety feature in Imessage. It employs on-device machine learning to analyze photo and video attachments and determine whether photos and videos appear to contain nude.
The grapheno operating system maintainer for posts shared on X provides a machine learning model that Safetycore does not provide client-side scans, primarily used in other applications to classify content I’ve repeated that it’s designed to be. Spam, scams, or malware.
“Classifying this kind of thing is not the same as trying to detect illegal content and report it to the service,” Grapheneos said. “It would violate people’s privacy in multiple ways, and there will still be some false positives. That’s not this and it can’t be used.”
Source link
