Google announced a set of new AI safety measures for India on Thursday at an event in New Delhi, saying these tools are meant to make AI usage more secure and help people protect themselves from misuse. The company noted that the updates include safeguards for financial applications, a new method to replace SMS-based OTP verification, and broader access to its SynthID watermark-detection technology.
In a blog post, Google explained that these steps aim to shield vulnerable users from online risks, strengthen privacy and cybersecurity for businesses, and ensure AI models remain fair and inclusive. The company said the rollouts were being made ahead of India’s AI Impact Summit in February 2026.
Google is collaborating with Google Pay, as well as financial apps Navi and Paytm, to test a new screen-sharing scam-alert system. On Android 11 and above, users will receive a warning if they open any of these apps while screen-sharing with an unknown person during a call, NDTV reported.
The firm is also developing what it calls Enhanced Phone Number Verification (ePNV), described as an Android-based authentication protocol meant to replace SMS OTPs. Google said this method uses a secure, SIM-based, consent-driven check designed to avoid the weaknesses associated with traditional OTP systems.
It added that the on-device Scam Detection tool powered by Gemini Nano has started reaching Pixel phones. This feature identifies possible scam calls locally on the device without recording audio, creating transcripts, or sending data to Google. Notably, this is turned off by default.
The company further pointed out that Google Play Protect has stopped more than 115 million attempts to install sideloaded apps that misuse sensitive permissions commonly exploited for financial fraud in India. Meanwhile, SynthID, Google’s AI watermarking and detection system, is being made available to researchers, academic institutions, and media publishers.