Google’s On-Device AI Scam Detection for Indian Users

Google has introduced a real-time, on-device scam detection system for Pixel users in India, aiming to reduce phone-based financial fraud through privacy-preserving AI. The launch marks a broader safety push as digital transactions expand across the country.

AI-Driven Call Analysis on Pixel Devices

The new feature uses Gemini Nano AI to analyse live phone conversations entirely on-device. It flags suspicious language patterns and warns users when unknown callers attempt deceptive tactics. No call recordings or transcripts are stored, ensuring user privacy while strengthening fraud resistance.

Protection During High-Risk Digital Interactions

A screen-sharing alert triggers when users open payment apps such as Google Pay, Paytm or Navi during calls with unfamiliar numbers. The prompt allows users to end both the call and the screen-share instantly, reducing the risk of persuasive fraud attempts.

Education Programs for Safer Digital Environments

Google is expanding digital literacy initiatives to support teachers, parents, and young users. The Learn and Explore Online programme begins this December, supplying tools for age-appropriate online engagement. The company continues to build on earlier efforts that have trained thousands of educators across India.

Exam Oriented Facts

• The scam detection feature works fully on-device using Gemini Nano AI.
• Screen-sharing alerts activate when payment apps are opened during unknown calls.
• Google’s Learn and Explore Online initiative launches in India in December 2025.
• CyberPeace Foundation received funding to build AI-based cyber-defence tools.

Enhanced Verification and Content Authenticity Measures

Google is developing SIM-based Enhanced Phone Number Verification to reduce dependence on SMS one-time passwords, improving protection against financial fraud. A verified list of legitimate digital lending apps, published with the Reserve Bank of India, helps users avoid unregulated operators. The company is also widening access to SynthID watermarking technology to assist researchers and media organisations in detecting AI-generated images and audio.