As India seeks to leverage artificial intelligence to boost its economic growth, establishing a secure AI framework is essential for this goal. Essentially, safety serves as the backbone for transformative AI rather than a supplemental feature. With an increasing number of Indians gaining internet access each month, our capacity to tap into the exponential benefits of AI is wholly reliant on how secure individuals feel while navigating online.
Fraud schemes are becoming more advanced in India, driven by organized groups that employ methods such as digital arrests, screen-sharing scams, and voice replication to target significant financial dealings. These attacks diminish confidence—not just in personal judgment but also in the internet as a whole. The only viable response is a defensive system that can outpace fraudsters, persistently detect threats, and is integrated directly into the technology that people use every day.
As we approach the forthcoming AI Impact Summit 2026 in New Delhi, we have provided updates on Google’s comprehensive strategy to utilize AI to safeguard at-risk users from online dangers, develop strong privacy and cybersecurity solutions for businesses, and create AI models that are inclusive, fair, and representative. This reflects how we are transforming the concept of safety as fundamental infrastructure into tangible protections for individuals and companies in India.
Enhancements to product safety
Our devices play a vital role in our digital experiences, making it imperative to incorporate strong safeguards directly into the products and services we rely on every day.
- Real-Time Scam Detection in phone conversations: The Scam Detection feature, developed by Gemini Nano and implemented on Pixel devices, examines calls instantly and identifies possible scams directly on the device, ensuring no audio or transcripts are recorded, nor is data sent to Google. This feature is deactivated by default, is applicable only to calls from unfamiliar numbers (excluding saved contacts), emits a beep to alert callers, and can be disabled by the user whenever they choose.
- Security for Financial Applications: In collaboration with financial applications Google Pay, Navi, and PayTM, we are testing a new feature in India to address screen-sharing scams. For devices operating on Android 11 and above, a noticeable alert will appear if a user accesses one of these applications during a screen-sharing call with an unknown individual. This feature offers a simple option to disconnect the call and cease screen sharing, safeguarding users against potential scams.
- Google Play Protect has efficiently thwarted more than 115 million attempts to install sideloaded applications that request sensitive permissions often exploited for financial scams in India. Additionally, Google Pay’s mechanism issues over 1 million alerts each week regarding fraudulent transactions, actively securing the foundation of our digital economy.
- Systemic Safeguarding: We are leading the way in Enhanced Phone Number Verification (ePNV), a novel security framework for Android that substitutes SMS OTP methods with a secure, agreed-upon, SIM-dependent verification to enhance login security standards.
- SynthID Collaborations: We are offering preliminary access to SynthID Detector and API, which is Google’s technology for watermarking that discerns artificially created content, to select partners including educational institutions, researchers, and media organizations like Jagran, PTI, and India Today, among others.
