Google puts kids, teens and elderly at the centre of Global AI safety

At a recent event titled ‘Safe and Trusted AI’ in New Delhi, Google emphasized the necessity of embedding safety as a core component of Artificial Intelligence (AI) to harness its transformative potential. This event, which precedes the anticipated AI Impact Summit 2026, underscored the company's commitment to protecting children, teenagers, and the elderly as integral elements of its AI safety initiatives. Evan Kotsovinos, Google’s Vice President for Privacy, Safety, and Security, articulated that a robust safety framework is essential for AI to confidently penetrate critical sectors such as finance, healthcare, and education, ultimately fostering economic growth. He noted that with India’s rapidly expanding digital economy, Google is developing AI solutions that prioritize user trust during the nation’s AI transition. Kotsovinos remarked that AI-driven defenses implemented on devices provide a significant edge against malicious actors targeting individuals and essential public infrastructure. He views India as a testing ground for dependable AI safety, predicting that the journey towards equitable AI for the Global South will be spearheaded from India. Preeti Lobana, Google India’s Country Manager, elaborated on the company’s comprehensive “360-degree” approach to safety. This strategy integrates protections both on devices and in the cloud, enhances digital literacy, invests in ongoing safety research, and fosters collaboration within the broader ecosystem. Lobana stressed that the goal is to create an environment where safety is inherently prioritized. Google’s initiatives include deploying protections specifically aimed at safeguarding children, teenagers, and the elderly across its platforms. The company is rolling out features that operate directly on devices to enhance privacy and efficiency. For example, a new feature powered by Gemini Nano analyzes calls in real-time to identify potential scams without recording audio or sharing data with Google. This feature is disabled by default and applies to calls from unknown numbers, allowing users to opt out anytime. Additional safeguards will alert users when they are at heightened risk, such as during screen-sharing sessions with unknown callers while using financial applications. This innovative approach aims to thwart scams targeting banking apps during live conversations. Moreover, Google is expanding access to its SynthID Detector and launching the SynthID text watermarking tool as part of its Responsible GenAI Toolkit. This technology assists academic and research partners in recognizing AI-generated content. Institutions like Jagran, PTI, and India Today have gained early access to this tool. In its dedication to enhancing digital literacy, Google plans to introduce its Learn and Explore Online (LEO) program in India in December 2025. This initiative focuses on training educators, practitioners, and parents to utilize parental controls and create age-appropriate online experiences. Google is also granting $200,000 from Google.org to the CyberPeace Foundation to bolster AI-driven cyber-defense capabilities and foster safer online learning environments for young users. Furthermore, Google is rolling out the “Sach Ke Sathi, DigiKavach for Seniors” program, which will provide in-person training to over 5,000 seniors across 30 cities. In collaboration with the Reserve Bank of India, it has published a verified list of authorized digital lending applications to help users differentiate trustworthy apps. In partnership with IIT Madras and the Centre for Responsible AI (CeRAI), Google is advancing research in AI safety tailored to India’s diverse linguistic and technological landscape. Notably, Google is also launching CodeMender, a security system designed to autonomously detect and rectify vulnerabilities in open-source software, alongside other Privacy Enhancing Technologies (PETs) such as Private AI Compute, Parfait, and VaultGemma. Lastly, the company is developing Enhanced Phone Number Verification (ePNV) to replace less secure SMS-based authentication methods with a SIM-based verification system to bolster sign-in security.