Intro: In a rapidly digitising India, how do we bridge rural gaps, make AI ethical and turn Digital Public Goods into global influence? In this exclusive conversation, Abhas Vyas, Digital Governance Specialist, reveals bold financing models, inclusive tech designs, and governance frameworks that can transform policy into real-world impact. Speaking to Mahima Sharma of Indiastat, his solutions span from broadband to AI to DPG diplomacy—offering a masterclass in making technology truly work for Bharat. Read the exclusive only at Socio-economic Voices this week.
MS: India launched National Broadband Mission 2.0 in January 2025 to fibre-connect 2.7 lakh villages and reach 100 Mbps speeds by 2030. What bold financing and partnerships will ensure equitable access across remote areas? How can private innovators and public bodies collaborate to operationalise meaningful digital inclusion?
AV: India's National Broadband Mission 2.0 (NBM 2.0), launched in January 2025, aims to extend optical fibre to 2.7 lakh villages. This is to ensure 100 Mbps minimum speeds by 2030. Achieving this scale requires blended financing i.e. combining Digital Bharat Nidhi grants, concessional multilateral loans and private equity, to de-risk low-revenue rural deployments where population density is low, Infrastructure-sharing policies as part of PM Gati Shakti initiatives can further lower both CAPEX and OPEX.
This would further require a meaningful digital inclusion strategy between public bodies and private innovators cutting across the Policy, Product Innovation, People and Capacity building and Performance. Governments play the foundational role by ensuring enabling policies, affordable spectrum, rapid RoW clearances and shared infrastructure rules, as seen in NBM 2.0’s national targets for fibre rollout, speed and uptime.
The private sector complements this by designing affordable, localised services in areas like health, education, finance and agriculture which are tailored to low-data, multilingual, mobile-first contexts. E-Sanjeevani Telemedicine, run by the Ministry of Health with Apollo Hospitals, has delivered over 16 crore remote consultations to rural and tribal populations, bridging specialist care gaps. The Common Services Centers (CSC) have partnered with the Open Network for Digital Commerce (ONDC) to enable rural citizens to shop online via the Grameen e-Store app. Together, such collaborations ensure digital connectivity evolves into genuine community empowerment.
MS: IndiaAI Mission now supports nearly 200 proposals to build indigenous LLMs and expanded compute capacity to 34,000 GPUs. How should funding, compute access and governance evolve to ensure that these foundational models serve public- and private-good use cases ethically and affordably?
AV: India’s IndiaAI Mission was launched in 2024 with a budget of Rs. 10,300 crore over five years. It has already built a high-end common compute facility supported through competitive bids and empanelment of cloud service providers, democratising access for startups, researchers and students.
To ensure India’s large language models (LLMs) and other foundational AI assets serve both public and private good ethically and affordably, the Mission has prioritised:
However, funding diversification and using challenge grants alongside flagship investments can unlock innovation in sectors like healthcare, agriculture and education. Shared infrastructure through platforms such as the IndiaAI Compute Portal ensures equitable, subsidised access to GPUs. Unlike many countries where AI infrastructure is controlled by large corporations, this initiative ensures that small players have an opportunity to innovate.
Furthermore, it is important to have strong governance frameworks for the LLMs including open-weight models for public-interest use, model audits and environmental/energy disclosures will drive responsible adoption. By combining inclusive funding, accessible infrastructure and robust governance, India can ensure its AI ecosystem is not just Made in India, but thoughtfully Made for Bharat.
MS: The government committed Rs. 500 crore in the 2025-26 budget for AI Centres of Excellence across healthcare, agriculture and sustainable cities, plus Microsoft and IndiaAI’s skilling program to train half a million people by 2026. What is the ROI in human capital and innovation output.
AV: The government has set aside Rs. 500 crore in the 2025–26 Union Budget to create AI Centres of Excellence (CoEs) in healthcare, agriculture, and sustainable cities. Along with this, Microsoft and IndiaAI will work together to train 5 lakh people in AI skills by 2026. This is a clear push to turn AI policy into real economic and social benefits.
The return on this investment in people can be seen in three ways —
Right now, India’s digital economy makes up 11.74% of the GDP (Rs. 31.6 lakh crore) and is expected to grow to 13.42% by 2025. If even 60% of those 5 lakh trainees get AI-related jobs, the value added could be huge, especially in fast-growing sectors like healthtech and agritech.
Since 2010, India has filed more than 86,000 AI-related patents, making up over 25% of all technology patents in the country. From 2021 to 2025, AI patent filings are already 7 times higher than in the 2010–15 period, showing just how fast innovation is picking up. CoEs can help speed this up further, creating smart solutions in farming and healthcare that save both time and money.
To see the real impact at the grassroots, we’ll need to track local-level results. This includes the number of villages or blocks using AI-driven solutions from CoEs, how many people actively use these platforms each month at the Panchayat or block level, and whether services are available in local languages and on low-end devices. We can also look at how productivity changes — for example, the percentage increase in crop yield from AI-based farming advice — and count how many local youth have been trained in AI skills.
If these investments are tied to such clear, measurable results, the CoEs can become real drivers of equity and growth in India’s AI economy.
MS: India co-hosted the AI Action Summit in Paris and signed the global declaration on inclusive and sustainable AI alongside 57 countries. How can India transform this commitment into enforceable norms—especially around digital equity, energy impact and fair labor implications?
AV: At the AI Action Summit in Paris (2025), India joined 57 countries in signing the Global Declaration on Inclusive and Sustainable AI. This means India has promised to make AI fair, eco-friendly and good for everyone. To make these promises real, a few clear steps are needed.
First, India should write these principles into its own AI rules; for example, making it compulsory to test AI systems for bias in Indian languages, report the energy used to train large AI models and set up ways for workers to raise concerns if automation affects their jobs. The upcoming Digital India Act can be a good place to include these rules.
Second, public funding should come with conditions i.e if a startup or research project gets government money, it should meet clear standards for fairness, accessibility and sustainability. For eg. EU’s AI act Article 43 ties market access to conformity assessments, a model India can adapt to its socio-economic priorities.
Third, India should set up independent AI review groups with experts from government, industry, universities and civil society to check how AI systems are being used and publish annual reports.
Finally, India can work with groups like the Global Partnership on AI (GPAI) to create global trust marks or certificates for ethical AI, something along the lines of ISO, so our AI exports are both competitive and responsible.
MS: India is a global poster child for DPGs like Aadhaar, UPI and DigiLocker. But now that these systems are being exported (e.g. to Africa), how can India monetise or diplomatically leverage DPG diplomacy without undermining openness?
AV: India can monetise and diplomatically leverage its Digital Public Goods (Aadhaar, UPI, DigiLocker, India Stack) while preserving openness by adopting a “services-around-an-open-core” model.
First, keep the core APIs and standards open-source so partners can adopt them freely. But then, do charge for value-added services i.e customisation, integration, hosting, SLA-backed cloud deployments, training, certification and long-term support. NPCI’s international work shows the model in action: India signs MoUs and builds payments rails abroad while earning implementation and support fees (e.g., deals in Peru, Namibia, Trinidad & Tobago).
Second, India can create India-Stack Alliances that bundle technical transfer with capacity building and policy advice (data-protection, interoperability), financed through low-cost development loans or concessional packages from multilateral partners.
Finally, use DPG diplomacy as part of broader development partnerships (including technical assistance + trade) to build long-term influence while encouraging reciprocal innovation. This keeps the stack open and wins both soft power and sustainable revenue.
MS: Despite UPI’s success, digital payment frauds are rising, especially among the elderly, which BETTER accountability, audit or grievance redressal mechanisms can strengthen trust in DPGs without hampering ease of use? Explain with examples please.
AV: Digital public infrastructures (DPIs) like UPI, Aadhaar-enabled Payment System (AePS) and FASTag have transformed access to service. But frauds, particularly against elderly or first-time users are on the rise. The challenge is to strengthen trust without slowing down or complicating use.
Real-time risk detection can be embedded into the DPI layer. For example, Singapore’s PayNow pauses suspicious transactions for verification, while India’s FASTag auto-flags abnormal toll usage, UPI platforms like Paytm’s “Pi” and PhonePe’s “Guardrails” watch transactions closely and flags to the users to confirm if something looks unusual before sending money.
Shared liability rules can make service providers jointly responsible for quick redressal. The UK’s “Authorised Push Payment” system does this well by helping victims get their money back fast. India is still building similar protections but is moving in that direction.
Special safety flags for elderly users' accounts can trigger real-time checks. For instance, if an elderly person tries a risky payment, a trusted family member could get an alert or call to verify it.
A “safe mode” feature could limit large payments, block sending money to new contacts for 24 hours or allow transfers only to trusted people.
Awareness campaigns through local groups like Resident Welfare Associations (RWAs), pension offices and Common Service Centers (CSCs) can warn seniors that no government official asks for money over the phone. This helps prevent scams where fraudsters pretend to be police or officials.
MS: With the government promoting platforms like ‘NITI Aayog’s Aspirational Blocks Programme’ and ‘Digital Health Mission’, how can technology be tailored to local capacities — especially in health, skilling and agriculture — without replicating urban biases?
AV: Government programmes like the Aspirational Blocks Programme and the Ayushman Bharat Digital Mission aim to take technology deep into rural areas. But if tools are designed only for urban infrastructure—fast internet, high-end smartphones—they risk leaving rural users behind.
First, design for low-resource environments. Applications should work offline or with patchy 2G/3G networks. For example, the Kisan e-Mitra chatbot delivers crop advice via SMS and IVR, not just apps. In Healthcare, for example, UNICEF’s RapidPro IVRS tool provides maternal health information to mothers via simple phone calls, ensuring vital advice reaches women without smartphones or internet.
Second, use local languages and context. Platforms like Bhashini enable real-time translation so that farmers, ASHA workers and local entrepreneurs can access content in their native tongue.
Third, integrate with existing community systems. In health, the e-Sanjeevani telemedicine platform is embedded in PHCs and uses ASHA workers to guide patients through consultations. In skilling, the Skill India Digital platform partners with CSCs to reach youth in smaller towns.
Fourth, train local champions i.e teachers, panchayat members, SHG leaders, so they can guide others. This builds trust and adoption faster than centralised campaigns. The Ministry of Youth Affairs and Sports, through Digital India Corporation, introduced a WhatsApp chatbot linked to the My Bharat portal, a platform designed to empower youth with learning, mentoring, volunteering and upskilling opportunities.
Finally, measure adoption, not just rollout; track who is using the service, how often and for what outcomes, to ensure benefits reach rural and marginalised groups.
About Abhas Vyas
Mr Vyas holds an MBA from IIT Kanpur (2015–2017) and a B.Tech in Mechatronics. His professional journey spans roles at NITI Aayog and leading global consulting, driving digital transformation across diverse sectors. He has led national-level initiatives for the adoption of Digital Public Goods (DPGs) to enhance grassroots communication, land governance, agritech solutions and GIS-based decision support systems. His work includes fostering innovation and enabling technology-driven governance. With a strong foundation in strategy and execution, I specialise in leveraging emerging technologies to create scalable, inclusive and sustainable digital ecosystems for public service delivery.
About the Interviewer
Mahima Sharma is an Independent Senior Journalist based in Delhi NCR with a career spanning TV, Print, and Online Journalism since 2005. She has played key roles at several media houses including roles at CNN-News18, ANI, Voice of India, and Hindustan Times.
Founder & Editor of The Think Pot, she is also a recipient of the REX Karmaveer Chakra (Gold & Silver) by iCONGO in association with the United Nations. Since March 2022, she has served as an Entrepreneurship Education Mentor at Women Will, a Google-backed program in collaboration with SHEROES. Mahima can be reached at media@indiastat.com
Disclaimer : The facts & statistics, the work profile details of the protagonist and the opinions appearing in the answers do not reflect the views of Indiastat or the Journalist. Indiastat or the Journalist do not hold any responsibility or liability for the same.
"Inclusive AI Needs Open Data, Subsidised Computing and Governance"... Read more
