Google develops ‘safe and trusted’ AI to protect vulnerable users in India

Google has announced that it is developing safe and trusted Artificial Intelligence as part of a broader push to safeguard vulnerable users in India, stating that safety must remain the foundation for transformative AI. The company outlined a wide-ranging set of initiatives ahead of the AI Impact Summit 2026 in New Delhi, emphasising protection from advanced scams, strengthening cybersecurity, and building inclusive AI models tailored for India and the Global South.

The India-AI Impact Summit 2026, set for 19–20 February in New Delhi, will be the first global AI summit hosted in the Global South. It was announced by India at the France AI Action Summit and will bring together international leaders to discuss responsible AI development.

Citing the rapid rise of digital arrest scams, screen-sharing fraud, and voice cloning, Google said its strategy focuses on embedding safeguards directly into everyday technology and ensuring protections act faster than scammers. The company is rolling out real-time scam detection on Pixel phones, powered by Gemini Nano, which analyses suspicious calls on-device without recording audio. A new pilot with Google Pay, Navi and PayTM will warn users when they open financial apps while screen-sharing with an unfamiliar contact, offering a one-tap option to exit safely.

Google Play Protect has blocked more than 115 million attempted installations of high-risk sideloaded apps in India, while Google Pay issues more than 1 million weekly fraud warnings. To improve account security, Google is introducing Enhanced Phone Number Verification, which replaces SMS OTPs with a secure SIM-based authentication process.

To address the growing threat of deepfakes, Google is expanding access to SynthID, its AI watermarking technology, to partners including PTI, Jagran and India Today. In cybersecurity, the company has introduced CodeMender, an AI agent that autonomously identifies and patches software vulnerabilities.

Google is also investing heavily in digital literacy programmes. Initiatives such as LEO, Super Searchers and the senior-focused DigiKavach aim to equip millions of people with the skills needed to recognise online risks. Through the Google.org APAC Digital Futures Fund, the CyberPeace Foundation will receive USD 200,000 to strengthen AI-driven cyber-defence tools.

The company said India’s scale, linguistic diversity and rapid digital growth make it a critical market for developing AI tools suited to the Global South. Google is expanding its partnerships with IIT Madras and CeRAI to advance AI safety benchmarks, datasets and governance frameworks.