Android security 2015 annual report: Google checks 6B+ installed apps per day for malware, and fewer than 0.15% of devices using only Google Play have malware
Android Security 2015 Annual Report — Posted by Adrian Ludwig, Lead Engineer, Android Security
Context & Ripple Effects
This annual report is Google's follow-up to its claim that it had cut Android malware in half in 2014, and it extends that argument with two new numbers: 6B+ installed-app scans per day and a sub-0.15% infection rate for devices that install exclusively from Google Play.
The framing matters because independent trackers kept finding most new Android malware concentrated in third-party app stores — a split G Data's hourly counts would quantify in 2017. The report is effectively a defense of Play's walled garden as the security boundary.
First-order effects
- Device makers and carriers get an official benchmark they can cite: staying inside Google Play means sub-0.15% malware exposure, making sideloading and alternative stores the visible risk factor on every Android deployment.
Second-order effects
- Independent security firms like G Data respond by publishing their own threat tallies that highlight exactly where Google's number doesn't reach — third-party storefronts — turning measurement itself into competitive pressure against Play's closed model.
Third-order effects
- The scan-first approach evolves into gatekeeping at the source: by 2017 Google reports 60.3% of potentially harmful apps detected via machine learning, and by 2024 it is rejecting millions of apps and blocking hundreds of thousands of developer accounts outright — though researchers later found Play Protect identifying under 70% of malicious apps in one lab test, showing the gap between scale claims and detection accuracy remains contested.
The trend: Android security is migrating from post-installation scanning toward machine-learning review and upstream enforcement at the app-store gate, with Google Play positioned as the dividing line between protected and exposed devices.