New AI-Powered Protections on Android Stop Scams and Malware in Real-Time

Google has announced the launch of two new AI-powered security features for Android devices, aiming to enhance real-time protection against scams and malicious apps while prioritizing user privacy. The new safeguards, Scam Detection in Phone and Google Play Protect’s live threat detection bring intelligent, on-device protection to Pixel users first, with wider Android availability anticipated soon.

Detecting scams in real-time

The Scam Detection feature, available in beta for English-speaking users in the U.S. with Pixel 6 or newer devices, leverages Google’s advanced on-device AI model, Gemini Nano, to detect potential scam calls in real-time. Using AI to identify patterns typical of scam conversations, such as urgent fund transfer requests allegedly from banks, the system alerts users with audio, visual, and haptic warnings.

Designed with privacy as a central focus, Scam Detection processes and analyzes calls entirely on-device, meaning no conversation data is stored or transmitted externally. The feature is disabled by default, allowing users to activate it as needed for future calls and maintain full control over their call data and privacy.

Blocking stalkerware apps

In tandem, Google Play Protect now includes live threat detection, which aims to enhance defense against harmful apps, focusing initially on identifying and alerting users about stalkerware — a type of software that secretly monitors personal or sensitive data. Live threat detection works by assessing behavioral signals like an app’s use of sensitive permissions or unusual interactions with other apps and services, particularly focusing on applications that hide suspicious behaviors or stay dormant before activation.

If Play Protect detects a potentially harmful app, users receive a real-time alert, giving them an opportunity to act quickly to protect their devices. Like Scam Detection, Google Play Protect’s new capabilities are powered by on-device processing, using Google’s Private Compute Core to ensure that no user data is sent to Google servers, thereby safeguarding user privacy.

Google’s ongoing push to improve Android security is underscored by its commitment to using on-device AI for processing and privacy protection. According to Google, scam calls are increasingly prevalent and more sophisticated, with scammers stealing over $1 trillion globally each year. Scam Detection and Play Protect’s live threat detection are intended to mitigate these risks by empowering users to spot potential threats before they can cause harm.

These real-time protection features are currently rolling out to Pixel 6 and newer devices, with plans to expand availability across other Android devices in the coming months. Users interested in trying out the new Scam Detection beta can enable it in the Phone by Google app and provide feedback via the app’s Help & Feedback section.

Google says it will continue to expand the scope of these features, making them available to more device models and broadening Play Protect’s detection capabilities for other types of harmful apps in the future.