Google introduces AI safety protection tools

Google said safety must function as ‘core infrastructure’ for artificial intelligence to deliver meaningful economic and social impact, setting the tone for its ‘Safe and Trusted AI’ event held in the capital ahead of the AI Impact Summit 2026.

The company said protection for children, teens and seniors remains a major priority. Many of the new features run on-device, a design the company says enhances privacy and provides real-time safeguards.

Among the product features announced:

• Real-time scam detection on calls: Using Gemini Nano, the tool analyses calls on-device and flags possible scams. The feature is off by default and applies only to unknown numbers. It does not record audio or share data with Google.
• Screen-sharing fraud warnings: A pilot alerts users if they are screen-sharing and open Google Pay, Navi, or Paytm during a call with an unknown number.
• SynthID Detector: Google is expanding access to its AI-generated content watermarking tools, providing early access to partners including Jagran, PTI and India Today.

Evan Kotsovinos, Vice President, Privacy, Safety and Security, Google, said: “Safety is the infrastructure for transformational AI. It is what allows AI to move with confidence into finance, healthcare, education and be the foundation for economic growth. In India, where the digital economy is booming, we are building AI systems that are designed to keep user trust intact as India transitions to AI. When AI-powered defenses run on the device, it gives us an unprecedented advantage over bad actors seeking to target individuals or critical public infrastructure. India’s scale and multiplicity make it the world’s proving ground for reliable AI safety. The path to equitable AI for the entire Global South will be led from India, and we are committed to supporting that journey.”

Preeti Lobana, Country Manager, Google India, said: “At Google, we are taking a 360 degree approach to safety - combining on-product and on-Cloud protections, digital literacy to empower users with knowledge and awareness, investing in continuous safety research so that we keep updating our defences along with the shifting goalpost and finally, joining hands with the ecosystem to ensure the entire chain gets stronger and smarter. The idea is to build an ecosystem where safety is embedded by default.”

The company also noted that it will launch its Learn and Explore Online (LEO) program in India in December 2025 to train teachers, practitioners and parents on online safety and parental tools.

It mentioned some of its ongoing campaigns:

• ‘Sach Ke Sathi, DigiKavach for Seniors’ with Jagran, which provides in-person training to more than 5,000 seniors in 30 cities and reaches about one million people indirectly.
• The Super Searchers information literacy program, which has trained more than 17,000 teachers and 10,000 students this year, reaching more than one million people through a train-the-trainer model.
• A $200,000 grant to the CyberPeace Foundation to support AI-driven cyber-defense tools, digital safety initiatives for children and teens, and capacity-building aligned with the IndiaAI Mission.

Google said it is also working with regulators, academia and civil society to strengthen digital safety and AI governance.