Google on Thursday announced a set of AI-powered security initiatives aimed at strengthening the digital safety framework, with the highlight being a scam-detection feature on Pixel phones powered by Gemini Nano. The tool uses on-device intelligence to analyse live call patterns and alert users to suspected fraud attempts without uploading or transcribing the conversation to the cloud. Google said the model was developed in partnership with fintech players including Google Play, Navi and Paytm, signalling a deeper integration of safety features into the financial ecosystem.
The announcements are aligned with the company’s privacy enhancing technologies (PETs) approach. This includes techniques such as federated learning, homomorphic encryption and differential privacy, elements which according to Meity officials are in line with provisions of the Digital Personal Data Protection Act (DPDP).
What else is new?
Alongside scam-detection, Google introduced an Android-based security protocol called enhanced phone number verification (ePNV), designed to offer a secure alternative to SMS-based one-time passwords. The system verifies a device-linked phone number directly through mobile operators using cryptographic checks. By cutting dependence on SMS OTPs, Google hopes to curb phishing attacks that exploit text-based authentication.
The company said that Android’s AI defences block around 10 billion malicious messages, spam and scam calls globally every month, with nearly 2 billion of those interventions occurring in India through lightweight models running locally on smartphones. Google Pay alone issues nearly a million warnings per week for potentially fraudulent transactions.
Google’s plan to tackle deepfakes
To tackle misinformation and deepfakes, Google showcased its SynthID watermarking technology, already embedded into more than 10 billion pieces of AI-generated content across images, videos and audio. The watermark remains imperceptible to users but allows platforms to identify synthetic media without degrading the viewing experience.
The company also reaffirmed long-term localisation plans through the Google Safety Engineering Centre in India and collaborations with IIT Madras and the Centre for Responsible AI to develop language-agnostic safety benchmarks and expand AI talent pipelines.
