Responsible AI for Bharat: Building Trust, Safety, and Global Leadership

Abstract

The session explored how India can evolve from a consumer of frontier AI models to a creator of responsible, people‑centric AI systems. Panelists argued for frugal, indigenous innovation, robust governance‑as‑infrastructure, and multilingual fairness as the pillars that will enable AI to serve the country’s 1.4 billion‑strong, linguistically diverse population while positioning India as a global leader for the Global South. A brief address by Minister Jitin Prasada underscored government commitment to inclusive, scalable AI and invited all stakeholders to contribute to a “human‑in‑the‑center” AI future.

Detailed Summary

• Amlan Mohanty (moderator) opened by congratulating the organizers for branding the summit around impact rather than merely declarations. He highlighted the inclusive ambition of the “AI Impact Summit” – to benefit all 8 billion people on the planet, not just a select few.
• He reminded the audience that India is the world’s third‑largest economy by PPP and emphasized that future economic size—not present GDP—will determine AI leadership. By extrapolating Indian GDP to European and American standards, he projected a potential $70–$170 trillion economy, surpassing today’s global GDP of roughly $120 trillion.

2. Frugal Innovation & Indigenous AI Foundations

• Suresh Yadav (panelist) argued that Western AI models are built for specific data contexts and that their sheer scale of parameters (billions, trillions) does not suit India’s heterogeneous linguistic, geographic, and literacy landscape.

• He cited India’s Digital Payments Infrastructure (DPI) – the Aadhaar‑linked UPI ecosystem – as a proof point that frugal, large‑scale digital infrastructure can drive rapid development.

• The panel highlighted three Indian assets that can be recombined into an indigenous, India‑driven AI model:

  1. Open data sets released by government agencies.
  2. Bhashini – a multilingual language‑model initiative.
  3. The nation’s massive higher‑education system (≈ 40 million students).

• Recommendations included converting the higher‑education talent pool from “users” to “creators” and leveraging voice‑driven, language‑agnostic interfaces to bring non‑English speakers into the AI creation loop.

3. Governance as Architecture – Trust‑by‑Design

• Navrina Singh (referred to as “Naveena”) posed two questions to Suresh Yadav about:

  1. Why robust governance should be treated as infrastructure, not an after‑thought.
  2. Whether India should develop its own governance‑risk‑compliance (GRC) frameworks to avoid outsourcing high compliance costs.

• Suresh Yadav responded that India currently consumes frontier models, making it dependent on opaque third‑party foundations. He introduced the concept of a “software bill of materials” for AI, urging builders to understand the provenance of training data, model architecture, and downstream impact before trusting a system.

• He stressed three pillars for trustworthy AI:

  1. Embedded governance from the outset (design‑time trust, not post‑deployment checks).
  2. Supply‑chain transparency – mapping every vendor that contributes to a chatbot or decision‑support system.
  3. Context‑specific evaluation benchmarks – fairness and reliability metrics calibrated for India’s demographic realities rather than imported Western standards.

• Suresh concluded that without these, India would miss the “AI arms race” and lose the chance to set global standards for trustworthy AI.

4. Multilingual Benchmarks, Data Infrastructure & Human‑Centric Evaluation

• Pratyush Kumar was invited to elaborate on multilingual evaluation. He described the need for benchmark datasets that reflect regional accents, vocabularies, and cultural contexts.

• He argued that “humanizing” AI – making models behave like collaborative co‑workers – must incorporate subjective user experience: voice‑matching, local idioms, and scenario‑specific vocabulary.

• Pratyush highlighted ongoing efforts to create evaluation pipelines that ingest diverse Indian language corpora, akin to how smartphone hardware is stress‑tested for real‑world usage.

• Santosh Vishwanathan framed AI as a nation‑building tool: just as UPI unlocked financial inclusion for micro‑enterprises, AI‑driven solutions for agriculture, health, and education can empower farmers and teachers at the village level. He argued for small, domain‑specific models (e.g., crop‑disease detectors) that run on low‑cost devices, rather than relying solely on massive global models.

5. Minister Jitin Prasada’s Keynote – Governmental Commitment & Vision

• The moderator announced the arrival of Minister of State for Commerce & Industry and Electronics & IT, Jitin Prasada.

• Minister Prasada welcomed participants, called the summit “historic”, and reiterated the government’s agile, non‑over‑regulatory stance aimed at fostering AI innovation.

• Key points from his speech:

  • Broadening GPU access at low cost for startups and research institutions.
  • Emphasis on scalable, impact‑focused AI models that address sector‑specific challenges (agriculture, health, education).
  • Alignment with Prime Minister Modi’s vision of democratizing technology beyond mere access—ensuring solutions are usable and affordable for the bottom‑up.
  • Commitment to positioning India as the voice of the Global South in international AI standard‑setting, guaranteeing that “have‑nots” are not left behind.

• The minister invited the panelists to share concrete policy needs; the moderator noted that access to compute, capital, talent, and market were already being addressed, while trust and governance were the remaining gaps.

6. Panel Reflections on Education, Skilling & Talent Development

• Santosh Vishwanathan compared AI’s potential impact on education to UPI’s disruption of payments. He advocated for an AI‑enabled, learner‑centric education system that moves away from static curricula to personalized pathways, leveraging AI to adapt to each student’s strengths and language.
• Suresh Yadav added that the government is contemplating curriculum changes but warned that without a holistic, nation‑wide upskilling strategy, India risks losing its competitive edge within the next decade.

7. Global‑South Perspective – AI as a Nation‑Building Lever

• Rish (Panelist – “Rish”) broadened the discussion to other developing nations, warning that Western AI standards may become barriers if they are imposed without participation from the Global South.
• He highlighted the need for inclusive standard‑setting, ensuring that new governance frameworks are co‑created and not used to block indigenous solutions.
• Rish argued that small‑model, frugal AI (e.g., mobile‑based disease‑diagnosis tools) and large‑model collaboration are not mutually exclusive; both should coexist to serve diverse development goals.

8. Closing Reflections & Outlook for the Summit

• With less than ten minutes remaining, the moderator asked each panelist to state their hope for the summit’s outcome:

  • Navrina Singh – A concrete, trust‑centric AI governance framework that can serve as a global blueprint, widespread AI literacy, and increased VC funding for trustworthy‑AI startups.
  • Pratyush Kumar – A balanced stance of competition + collaboration with frontier model owners, ensuring India’s narrative of “AI for everyone” (Sarvajana Hitaya, Sarvajana Sukhaya) shapes global discourse.
  • Santosh Vishwanathan – Validation that frugal innovation can coexist with high‑impact AI, proving that efficient, low‑resource models are a global benefit.
  • Suresh Yadav – Confirmation that the Global South will not just follow but lead the development of responsible AI standards, with “human‑in‑the‑center” as the rallying principle.

• The session wrapped with a round of applause, gratitude to organizers (Primus Partners, Sunita, Indranuj, Sakshi), and a reminder that the next panel would start immediately after a short break.

Key Takeaways

• India’s future AI leadership hinges on scaling from a consumer to a creator; leveraging its massive data sets, multilingual initiatives (Bhashini), and the world’s largest higher‑education ecosystem.
• Frugal innovation is a strategic advantage – the DPI/UPI experience shows that low‑cost, high‑reach infrastructure can drive inclusive growth; similar models should be applied to AI (e.g., region‑specific small models).
• Governance must be built into AI systems from day one; treating trust as infrastructure rather than a post‑hoc compliance exercise. Transparency of the AI supply chain is essential.
• Western fairness benchmarks do not automatically apply to India; new, multilingual, culturally‑aware evaluation metrics are required to measure reliability, bias, and user experience.
• Government policy is already addressing compute access, talent, and market creation, but trust‑centric governance frameworks remain the critical missing piece.
• The Minister of State reaffirmed the government’s agile, non‑over‑regulatory stance, promising low‑cost GPU access and positioning India as the voice of the Global South in international AI standard‑setting.
• Education and skilling must shift from static curricula to AI‑enabled, personalized learning, turning millions of students into future AI builders rather than just users.
• Global‑South collaboration is essential; standards should be co‑created to avoid imposing barriers that would stifle indigenous AI development.
• Consensus emerged that “human‑in‑the‑center,” not merely “human‑in‑the‑loop,” should be the guiding principle for responsible AI—placing people’s welfare at the core of design, deployment, and evaluation.

These take‑aways capture the core messages, strategic recommendations, and forward‑looking commitments that surfaced during the “Responsible AI for Bharat” panel at the AI Impact Summit in Delhi.

See Also:

• democratizing-ai-resources-and-building-inclusive-ai-solutions-for-india
• building-resilient-sustainable-ai-infrastructure-for-people-planet-and-progress
• ai-innovators-exchange-accelerating-innovation-through-startup-and-industry-synergy
• preparing-to-monitor-the-impacts-of-agents-closing-the-global-assurance-divide-for-safe-and-trusted-ai
• safe-and-trusted-agentic-ai-building-accountability-and-inclusion-for-india-and-the-global-south
• scaling-ai-solutions-through-southsouth-collaboration
• ai-for-economic-growth-and-social-good-ai-for-all-driving-economic-advancement-and-societal-well-being
• enterprise-grade-responsible-ai-for-india-and-the-global-south