What Happened
- PM Modi unveiled the "MANAV Vision" for human-centric AI at the summit's Day 4, urging global leaders to work on a roadmap for responsible AI impact.
- Swiss President Guy Parmelin announced that Switzerland will host the next AI Summit in Geneva in 2027, with the UAE to host in 2028.
- Reliance Industries committed Rs 10 trillion to AI and digital infrastructure investment.
- India showcased indigenous AI models: Sarvam AI launched 30B and 105B parameter large language models; BharatGen introduced Param2, a 17B parameter multilingual model built with Nvidia.
- Google announced the America-India Connect initiative for new strategic fibre-optic routes between the US, India, and the Southern Hemisphere.
- Union Minister Vaishnaw announced 38,000 GPUs onboarded under the national compute platform, with 20,000 more planned.
Static Topic Bridges
AI Summit Progression — From Bletchley to New Delhi to Geneva
The global AI governance summit series represents an emerging multilateral framework for AI regulation. The UK hosted the inaugural AI Safety Summit at Bletchley Park (November 2023), producing the Bletchley Declaration signed by 28 countries including China. France followed with the AI Action Summit in Paris (February 2025), broadening the agenda from safety to inclusive development and open-source AI. India's AI Impact Summit (February 2026) shifted focus to the Global South. Switzerland's 2027 Geneva summit will leverage the city's status as a hub of multilateral governance (home to WTO, WHO, UNHCR, ICRC, ITU).
- Bletchley Park AI Safety Summit (UK): November 2023 — 28 countries, Bletchley Declaration
- Paris AI Action Summit (France): February 2025 — broadened to inclusion and open-source
- India AI Impact Summit: February 16-20, 2026, New Delhi — 100+ countries, 20 heads of state
- Next summits: Geneva 2027 (Switzerland), UAE 2028
- India's summit pillars: People, Planet, and Progress
- Key attendees: French President Macron, Brazilian President Lula, Swiss President Parmelin, CEOs of Google, OpenAI, Anthropic
Connection to this news: The summit series is consolidating into an annual multilateral AI governance mechanism, with Switzerland's selection of Geneva reflecting the desire to anchor AI governance within established international organisations (ITU, WTO, WHO).
India's Sovereign AI Strategy — Indigenous Models and Compute Infrastructure
India's sovereign AI strategy centres on building domestic AI capabilities to reduce dependence on foreign technology companies for critical AI infrastructure. This includes indigenous large language models trained on Indian languages, subsidised national compute platforms, and policy frameworks that emphasise data sovereignty ("whose data, his right"). The IndiaAI Mission provides the institutional framework, with over 38,000 GPUs available through public-private partnership.
- IndiaAI Mission: national programme for compute, data, and capacity building
- Compute platform: 38,000 GPUs onboarded, 20,000 more planned — subsidised access for startups and academia
- Indigenous LLMs showcased at summit: Sarvam AI (30B and 105B parameter models), BharatGen Param2 (17B parameter, built with Nvidia)
- AIKosh: national AI dataset platform with 9,500+ datasets and 273 sectoral models
- National Supercomputing Mission: 40+ petaflop systems including AIRAWAT, PARAM Siddhi-AI
- India Semiconductor Mission: Rs 76,000 crore outlay, 10 projects approved
- Reliance commitment: Rs 10 trillion for AI and digital infrastructure
Connection to this news: The summit served as a launchpad for India's sovereign AI capabilities, demonstrating that Indian companies can build competitive multilingual models — a key argument underpinning India's call for democratised and inclusive AI.
Digital Public Infrastructure (DPI) — India's Global Proposition
India's Digital Public Infrastructure (DPI) model — comprising the India Stack (Aadhaar, UPI, DigiLocker, ONDC) — has emerged as a replicable template for developing countries. At the G20 under India's presidency (2023), India positioned DPI as a framework for inclusive digital development. The AI summit extended this logic to AI infrastructure, arguing that compute and AI models should be treated as public goods, not private monopolies.
- India Stack components: Aadhaar (biometric ID, 1.3+ billion enrolments), UPI (unified payments, 16+ billion monthly transactions), DigiLocker (document storage), ONDC (open commerce)
- UPI: launched 2016 by NPCI; daily transactions exceed 500 million
- G20 New Delhi (2023): DPI framework adopted in leaders' declaration
- DPI exported to: 10+ countries adopting India-inspired digital payment systems
- AI as public infrastructure: 38,000 GPUs as shared compute resource under IndiaAI Mission
- Google's America-India Connect: new fibre-optic routes for digital connectivity
Connection to this news: India's positioning of AI compute as shared national infrastructure (like UPI for payments) extends the DPI model into the AI domain, reinforcing the country's narrative of technology as an enabler of inclusion rather than a source of monopoly.
Key Facts & Data
- India AI Impact Summit 2026: February 16-20, New Delhi, 100+ countries, 20 heads of state
- MANAV Vision: Moral, Accountable, National sovereignty, Accessible, Valid
- Next AI Summits: Geneva 2027 (Switzerland), UAE 2028
- Reliance AI commitment: Rs 10 trillion
- IndiaAI compute: 38,000 GPUs onboarded, 20,000 more planned
- Indigenous LLMs: Sarvam AI (30B, 105B parameters), BharatGen Param2 (17B parameters)
- Google's America-India Connect: new fibre-optic routes
- Previous AI summits: Bletchley Park (2023), Paris (2025)
- Aadhaar enrolments: 1.3+ billion; UPI: 16+ billion monthly transactions