Google launches real time scam alerts and AI safety tools for Indian users

New Delhi: Google has laid out a new safety focused plan for India that puts scam protection, AI security and digital literacy at the centre of its work. The company shared the updates ahead of the AI Impact Summit 2026 in New Delhi, calling safety the foundation of India’s AI future. Evan Kotsovinos, Google’s vice president for privacy, safety and security, said safety is the infrastructure for transformational AI and not something added later. He also said the path to equitable AI for the Global South will be led from India.

The announcement comes at a time when scam cases in the country are growing fast. Many of these frauds now involve digital arrest threats, screen sharing traps and voice cloning. I have personally heard people around me talk about strange calls using AI voices, and it is easy to see how someone unfamiliar with technology might get tricked. Google said the only effective answer is a protection system faster than the scammer and built directly into devices people already use.

Google is rolling out a set of India focused protections on Android and Pixel devices. The most important one is Real Time Scam Detection on phone calls. Powered by Gemini Nano, the feature analyses calls on the device and flags suspicious patterns. It does not record audio or send data to Google. The company noted that it works only for unknown numbers, stays off by default and plays a beep to notify participants.

Google is also piloting a financial safety feature with apps like Google Pay, Navi and PayTM. If a user opens one of these apps while screen sharing with an unknown contact, the phone shows a prominent warning with an option to immediately end the call and stop screen sharing. This is aimed at tackling the growing problem of screen sharing scams, which have caused many high value losses across Indian cities.

Google Play Protect continues to block sideloaded apps with sensitive permissions that are frequently used for financial fraud. According to the company, more than 115 million installation attempts have been stopped in India. Google Pay also displays over 1 million warnings every week for risky transactions.

Google is looking at bigger structural fixes as well. Enhanced Phone Number Verification replaces SMS OTP with a secure SIM based check. This system aims to make sign ins more reliable and harder to hijack. Google also expanded access to SynthID Detector, its watermarking tool for synthetic content. Publishers like Jagran, PTI and India Today are using the tool to identify AI generated media.

On the cybersecurity side, Google is introducing CodeMender, a new code security agent that identifies zero day vulnerabilities and patches them automatically. It builds on earlier projects like OSS Fuzz, a free platform used by open source developers to find bugs in critical software.

The company is also training AI startups through programs like the Agentic AI Roadshow and the AI Agent Masterclass. These sessions help startups build secure AI agents using Google’s Secure AI Framework.

A large part of Google’s plan focuses on education. The LEO programme will launch in India in December 2025 and train teachers and parents on using parental tools and setting up safe online environments for children. The Super Searchers initiative has trained over 17,000 teachers and 10,000 students this year, and reached more than 1 million people. It teaches users how to evaluate information and identify AI generated content using tools like SynthID.

For senior citizens, the DigiKavach for Seniors programme runs in partnership with Jagran. It has provided in person training to over 5,000 seniors and reached more than 1 million end users across 25 cities. Google has also launched an interactive game called Be Scam Ready, which exposes users to real scam situations in a safe environment to improve their decision making.

Google.org is giving 1 million dollars to five institutions across APAC to study AI’s opportunities and challenges. In India, the CyberPeace Foundation will receive 200,000 dollars to build AI driven cyber defence tools for children, teens and vulnerable groups. It will also run hackathons to help developers create safety tools.

Google is expanding its work with IIT Madras and CeRAI to develop Hindi language AI safety benchmarks and support the AI Safety Commons through the AI Impact Summit. The company is also working with the Reserve Bank of India to maintain a public list of authorised digital lending apps.

India’s scale and multilingual reality continue to shape how AI safety models are built. Google said the frameworks developed here are creating a blueprint for the Global South.