Google has outlined a series of updates to strengthen AI safety in India, saying its focus has been on protecting vulnerable users as the country accelerates its digital growth. The company has said that safety has been positioned as core infrastructure for AI, and not an afterthought, as online risks have become more sophisticated.
It has noted that scams in India have grown more complex, with tactics such as digital arrest, voice cloning, and screen-sharing fraud targeting financial transactions. Against this backdrop, Google has said it has expanded its multi-layered approach to online safety, including real-time scam detection, privacy enhancements, cybersecurity tools, and digital literacy initiatives.
The company has shared that Real-Time Scam Detection on Pixel phones has begun rolling out, analysing calls on-device and flagging potential fraud without sending audio or transcripts to Google. It has added that a pilot to protect financial app users has been introduced with Google Pay, Navi and PayTM, where Android devices now warn users if they open these apps while screen-sharing with unknown contacts. Google Play Protect has also blocked over 115 million attempts to install sideloaded apps using sensitive permissions associated with financial fraud, while Google Pay has displayed more than 1 million weekly warnings for suspicious transactions.
Google has further highlighted progress in systemic protection efforts. It has said it is piloting Enhanced Phone Number Verification (ePNV) to replace SMS OTP flows with a secure SIM-based check. Early access to SynthID Detector and API has been provided to partners in academia, research and publishing to help identify synthetically generated content and strengthen transparency.
On the cybersecurity front, Google has launched CodeMender, a code security agent that autonomously identifies and patches vulnerabilities. It has added that initiatives such as the AI Agent Masterclass and Agentic AI Roadshows have trained early-stage startups on building secured AI agents using the Secure AI Framework (SAIF) 2.0. The company has also released privacy-enhancing tools and open-source libraries such as Private AI Compute, Parfait and VaultGemma.
Google has expanded its digital literacy and safety initiatives for young people, parents, low-income communities and seniors. It has said its LEO program will launch in India in December 2025 to train educators and families on parental tools. Through the Super Searchers initiative, over 17,000 teachers and 10,000 students have been trained directly this year, reaching more than 1 million end-users. Campaigns like DigiKavach and DigiKavach for Seniors have also reached millions through safety workshops, in-person sessions and awareness content. The company has additionally introduced the “Be Scam Ready” game to help users recognise fraud tactics.
Through Google.org’s APAC Digital Futures Fund, the CyberPeace Foundation in India has received support to develop AI-driven cyber-defense tools and strengthen responsible AI governance. Google has also worked with regulators and civil society groups, including collaborating with the RBI on a public list of authorised digital lending apps. It has expanded its engagement with IIT Madras and CeRAI across areas such as multilingual AI safety benchmarks, diverse datasets, and promoting the Secure AI Framework.
The company has said that India’s scale, multilingual environment and device diversity have created a unique testing ground for developing secure and equitable AI systems for the Global South.
“Safety is the infrastructure for transformational AI. We are building this foundation, and prioritizing the protection of vulnerable users. The path to equitable AI for the entire Global South will be led from India, and we are committed to supporting that journey.”
“SynthID is indeed one the most reliable detection tools out there.” — Pratyush Ranjan, Chief Editor, Digital Services, PTI
“By identifying AI-produced material with a single tap, it actually strengthens transparency and builds greater public trust in the integrity of information in today’s world.” — Bal Krishna, Senior Executive Editor, India Today
