Google on Wednesday introduced two new artificial intelligence (AI) security tools for Android devices. These tools aim to protect users from phone call-based scams and malicious apps by monitoring activity in real-time. The first is Scam Detection in Phones by Google, which monitors conversation patterns to determine if an incoming call may be a scam. The second is Google Play Protect real-time alerts that monitor the background activity of an app after it is installed to detect malicious apps.
The tech giant explained in detail about the two new security tools in its security blog post. Both of these features are being rolled out for Google Pixel 6 and newer models. The scam detection feature on the phone will initially be available only to those in the US who have signed up for the Google Beta program. The call-based security tool will only work on English language phone calls. Google Play Protect Live Alerts will also be available outside the US.
The scam detection feature differs from typical caller ID apps and services that use phone numbers and track calling behavior to determine whether a number is associated with scams. Instead, Google is using its on-device machine learning models to process a call’s conversation patterns in real time to determine whether an incoming call is a scam. On the Pixel 9 series, this will be done by the Gemini Nano.
Highlighting an example, the tech giant said that if a caller claims to be from the user’s bank and asks them to transfer money due to the breach, the AI model can process the audio information and send it to its database. can be used to confirm whether similar conversation patterns have been used. To deceive people.
Once it determines that the call could be a potential scam, the AI will provide an audio and haptic alert and show a visual warning. Google highlighted that the feature will be off by default, and users can either turn it on for all calls from the Phone app settings or turn it on for a particular call. The company claimed that none of the conversation audio or transcriptions are stored on the device, sent to Google servers or anywhere else, or retrieved after the call.
Google Play Protect Live Threat Detection feature
The second feature is part of Google Play Protect, a security tool that monitors the Play Store for malicious and harmful apps. With the AI-powered Live Threat Detection feature, Google’s AI models will monitor apps installed on eligible Android devices. If an app shows suspicious behavior or unnecessary interactions with other apps, the tool will issue a warning, alerting the user in real time.
Google claims that this AI tool will be able to detect apps that remain inactive for some time after installation to avoid suspicion. Furthermore, since users can detect harmful behavior in real-time, it will help them take timely action to prevent data theft. Notably, at launch, the tool will only focus on stalkerware and malicious apps that collect personal or sensitive user data without permission.