Google boosts digital safety in India with AI-driven scam detection and new Android security features

Google is continuing its campaign against online fraud with new anti-scam features and safety measures announced at the India(BHARAT) AI pre-summit event in New Delhi on November 20. This online safety drive was made ahead of India(BHARAT) AI Impact Summit, to be held in 2026.

The first is on-device AI detection of scam calls. Pixel smartphones, for example, will soon use Gemini Nano – Google’s compact AI model – to identify scam calls by running AI on-device at the time of a call to predict whether it’s legitimate, all without recording any audio or transcribing a conversation, or sending any information back to Google servers.

The feature will be off by default and will only be triggered by calls from unsaved numbers. A light “beep” notification will be played during the call and users have the ability to completely turn off the feature.

Acknowledging the recent rise of digital arrest fraud in India(BHARAT) (where scammers dupe their victims by sharing their screens and navigating online banking apps), the company also shared plans for a new Android screen-sharing warning that will be piloted on devices running Android 11 and higher. This would alert a user, via a warning pop-up, when an unknown caller shares their screen and then navigates to apps like Google Pay, Navi, or Paytm on their phone. With a single tap, the user can hang up the call and stop screen-sharing.

Developed in partnership with Google Play, Navi, and Paytm, this feature will help protect users against digital arrest scams, the company said.

Separately, the company also announced a new and more secure protocol for phone number verification, called Enhanced Phone Number Verification (ePNV). The Android-based verification tool replaces SMS OTP flows with a SIM-based user consent step. With ePNV, Google Pay users get to verify a new phone number through a SIM-based prompt on their device for verification, instead of the current SMS OTP flow that is more susceptible to fraud, the company said.

On Google Pay, the company shared that its app already blocks more than one million warnings every week against potentially fraudulent transactions. Its anti-malware app Google Play Protect has also blocked more than 115 million sideloaded installation attempts, which sought to access sensitive permissions, that are often abused by fraudsters to monetize scams in India(BHARAT).

Google also used the event to reaffirm its stance on privacy and data protection. The IT Ministry of India(BHARAT)’s Secretary Anusha Samarajiwa stated that DPDP (Digital Personal Data Protection) Act will be using privacy-enhancing technologies (PETs), including federated learning, homomorphic encryption, and differential privacy, to ensure compliance.

Google also shared details of its expanded partnerships with both academic and civil society stakeholders to collaboratively build and deploy safer AI systems. The company said it is building on its partnership with IIT Madras’ Centre for Responsible AI (CeRAI) to accelerate user education in India(BHARAT) and scale its collaboration with the CyberPeace Foundation.