Google unveils ‘ATL Saathi’ app for Indian teachers, expands local AI cloud and security offerings | Technology News

Google’s DeepMind unit in India has launched a new desktop web application that gives teachers access to a Gemini-powered assant trained on curriculum designed Atal Tinkering Labs under NITI Aayog’s Atal Innovation Mission (AIM).Known as ATL Saathi, the app can be used teachers to curate hands-on learning experiments for students. It is being rolled out to 100 schools across the country this year, with a long-term goal of reaching 10,000 schools in the future, Google said.The search giant also announced new initiatives for Indian developers alongside the launch of AI-powered educational tools, partnerships, and product expansions designed to help Indian AI companies work with Google while complying with data localisation requirements. These announcements were made Google at its annual I/O Connect India edition in Bengaluru on Tuesday, July 14.
Tech giants such as Google, Microsoft, and Amazon view India as a strategic growth market because of its large internet user base and vast pool of software developers. AI companies such as OpenAI and Anthropic have followed suit, establishing offices here to expand their presence in the country.
Google on Tuesday said that it paid out over Rs 5.3 lakh crore, up 28 per cent in 2024, to developers who are part of the Google Play and Android ecosystem in India, according to third-party research. These studies also found that over 91 per cent of surveyed Google Play app developers who used Google AI said that these tools directly or indirectly helped them generate more revenue, while 76 per cent of adults said they would be interested in using Google AI tools to ‘vibe code’ – the practice of building apps and other software simply prompting AI tools in natural language.
“As we drive the shift into the agentic era, where AI moves from answering queries to securely executing tasks, our focus is on providing the underlying infrastructure and guardrails the ecosystem needs to scale safely. Today, we are delivering just that across the board – from flexible on-premise cloud environments for highly regulated sectors, to open security protocols for developers, to localized tools for healthcare and classrooms. We want to ensure that the next wave of Indian innovation is secure, trusted, and built on locally relevant foundations,” Preeti Lobana, Country Manager, Google India, said in a statement.
Dr Manish Gupta, Senior Director for India and APAC at Google DeepMind, said, “The ultimate metric of AI progress isn’t just model parameters, but also in the positive transformation it enables. India is championing this as it adopts AI across every tier of the economy – from local merchants to national health initiatives.”Story continues below this ad

The India edition of Google I/O Connect 2026 saw over 1,500 developers in attendance. Several Indian startups such as Adya.AI, VideoSDK, Sivi, Superjoin, Knit, PolicyBazaar, Emergent, and RedBus also showcased their latest AI solutions powered Google’s AI models at the event.
Other major announcements
Google DeepMind on Tuesday said that it is working with institutions across India to scale AI Research Foundations, a free 56-hour programme for building and fine-tuning large language models (LLMs). It has partnered with Nasscom and IISC Bangalore along with AVPN, a network of social investors, to broaden accessibility to the education initiative.
To help Indian enterprises mitigate potential threats that might emerge with the misuse of agentic AI, Google said that it is bringing Sec-Gemini v3, its specialised cybersecurity agent, to trusted government and enterprise testers such as Flipkart. Sec-Gemini can reason across complex security data and help teams investigate incidents at machine speed, as per the company.
Google has also made its Gemini 3.5 Flash model accessible to Indian enterprises and startups via the Gemini Enterprise Agent Platform and the Gemini Enterprise app. The model has been made available in line with the country’s machine learning processing requirements.Story continues below this ad

To further ensure data localisation, Google said it will allow Indian enterprises, including regulated industries and the public sector, to run Gemini on Google Dributed Cloud from within Indian data centres. This means that Indian enterprises and public sector organisations can now build AI models entirely within their own perimeter, with supporting services also fully disconnected from the public internet.
Gemini Live, the company’s conversational AI agent, now supports 25 Indian languages and dialects, including Sanskrit, Bhojpuri, and Maithili.
In healthcare, access to Google’s multimodal MedGemma open models has been made available to researchers at AIIMS Delhi, who are using the models to develop new India-specific models for leprosy and sexual and reproductive health. Google said it is also open-sourcing CAPSEM, a secure runtime environment for an AI agent, for developers in the country. CAPSEM ensures that even if an agent is compromised or encounters a malicious prompt, the wider system remains fully protected.
The company is further bringing other open industry standards to India, including the Device Bound Session Credentials (DBSC) standard which makes stolen session cookies instantly useless to bad actors, and the Agents-to-Payments protocol, operating alongside the open Agent2Agent (A2A) protocol, that helps make authorised, low-value agent-led financial transactions (under $100) highly secure and accountable.Story continues below this ad
Researchers at IIT Delhi and IIT Madras are working with Google on research related to  agentic safety, including early threat detection and the development of next-generation ‘Guardian Agents’ that help mitigate institutional risks.