Google is adding additional checks to help users deal with scammers in India. This is courtesy of the rollout of its real-time scam-detection tool for Pixel smartphones, introducing an on-device system powered by its Gemini Nano model that analyses active calls and warns users of potentially fraudulent behaviour. The feature, unveiled at the company’s “Safe and Trusted AI” event, comes as phone-based scams and digital payment fraud continue to surge across the country. The introduction of on-device AI detection may give Indian smartphone users a stronger line of defence against rapidly evolving fraud techniques, reducing exposure during vulnerable moments such as financial transactions, screen sharing or unsolicited calls.

The tool evaluates calls from unknown numbers directly on the device and surfaces alerts when patterns resembling social-engineering attempts appear. The feature remains disabled by default and plays a notification tone audible to both callers when turned on. To secure privacy, the model neither records audio nor sends transcripts to its servers, and the feature can be stopped at any point during a call. The capability is initially available only on Pixel 9-series devices and later models, and it currently supports English-language interactions. Google said it intends to extend the feature to non-Pixel Android phones, although no schedule has been announced.

Alongside voice-call detection, Google introduced a screen-sharing alert designed to counter fraud involving UPI and banking apps. When a user is on a call, sharing their screen, and simultaneously opens financial applications such as Google Pay, Paytm or Navi, the device will display a warning and provide a one-tap option to end the call and halt screen sharing. The feature is available on devices running Android 11 or later and is part of a pilot program Google expects to expand to additional financial partners. The alert will also be available in Indian languages. The company said the protections run entirely on-device, limiting data exposure and reducing the risk of interception during sensitive interactions.

This comes at a time when India has witnessed a sharp rise in digital fraud as more consumers adopt online payments and smartphones become the primary gateway for banking and identity services. The Reserve Bank of India reported more than 13,500 digital-transaction fraud cases in 2024, accounting for over half of all bank-reported incidents, marking theft in the billions. The Ministry of Home Affairs estimated losses of approximately â‚č70 billion in the first five months of 2025 from online scams, noting that many victims do not file complaints. Authorities have warned repeatedly about phishing schemes, digital-arrest scams, impersonation calls and malicious apps seeking excessive permissions. Children, teenagers and senior citizens remain among the most targeted groups. Google said Android’s AI-driven defences now block around 10 billion malicious or spam messages, calls and scams globally each month, including nearly 2 billion detections originating from India. Google Pay surfaces more than a million warnings per week for transactions flagged as suspicious.

Google also highlighted progress on enhanced phone number verification, or ePNV, a device-based authentication standard intended to replace SMS one-time passwords. The feature performs SIM-based verification directly on the phone using cryptographic methods and works across Wi-Fi and mobile networks. The company said the system is being tested with partners in India as part of an effort to curb OTP-related fraud, which remains widespread. Separately, Google said Play Protect has blocked more than 115 million attempts this year to install sideloaded apps requesting high-risk permissions often used in predatory lending and impersonation scams. The company continues to face scrutiny from security researchers and law-enforcement agencies that have flagged fraudulent apps slipping through Play Store reviews.

The Tech Portal is published by Blue Box Media Private Limited. Our investors have no influence over our reporting. Read our full Ownership and Funding Disclosure →