Google Embeds Real-Time Scam Detection Directly on Android Devices, Keeping Analysis Local

Google has deployed on-device AI models across Android to detect fraud in calls and messages without transmitting data to servers. The system flags deepfake voice impersonation, suspicious SMS/MMS/RCS patterns, and delivers real-time warnings mid-conversation. Processing happens locally, reducing latency and eliminating privacy exposure—a significant shift toward embedded security at scale.