AI For People - Paris

Registration and Lunch

Inaugural Session

Session 1: From Paris to Delhi: How AI can Shape the World

AI is rapidly becoming the engine of productivity, strategic influence, and societal transformation. Against the backdrop of the India–France Year of Innovation, and with the Paris AI Summit in 2025 feeding directly into India’s hosting of the AI Impact Summit in 2026, this is an opportune moment to deepen holistic cooperation across universities, enterprises, researchers, and policymakers. A stronger France–India innovation corridor can translate summit momentum into concrete frameworks that drive growth, expand inclusion, and support sustainable development, especially across the Global South. This panel will examine how India and France can build an interoperable ecosystem of cooperation and how their shared leadership can shape an AI future that is secure, citizen-centric, and anchored in global prosperity.

• How can nations use AI to strengthen public institutions, improve service delivery, and build more responsive and citizen-centric governance systems?
• How can countries forge inclusive, cross-border partnerships that align innovation with development priorities and extend the benefits of AI to all societies?
• Could AI become the basis of a renewed, forward-looking multilateralism that prioritises cooperation, equity, and collective progress?
• What lessons from the Paris Summit can India carry forward to shape a more inclusive, interoperable, and globally trusted AI ecosystem?

Break

Session 2: Inclusion by Design: The Next Frontier of Global AI Innovation

AI is transforming economies and societies, yet its benefits remain unevenly distributed across languages, cultures, and geographies. As frontier innovation accelerates, the risk is that billions are left out, not because of a lack of capability, but because AI systems still reflect narrow datasets, unequal infrastructure, and concentrated technological power. The global challenge now is to make inclusion a strategic design choice: to build multilingual datasets, strengthen last-mile access, expand a[ordable compute, and create cross-regional innovation pipelines. This panel will explore how countries can embed inclusion at every stage of the AI lifecycle and ensure that AI becomes a force for empowerment, opportunity, and shared global prosperity.

• As AI rewrites global power structures, can inclusion become a strategic advantage rather than merely a development aspiration?
• What global governance frameworks are needed to ensure that fairness audits, diverse datasets, and inclusion-by-design become non-negotiable norms?
• Can India’s experience in scaling digital public infrastructure, multilingual datasets, and last-mile service delivery, serve as a global model for inclusive AI
  ecosystems?
• What role should media, education systems, and community organisations play in ensuring that AI literacy becomes a universal capability, not a privilege?

Session 3: Resilience First: Making AI Sustainable and Accessible

AI is advancing rapidly, but its foundations depend on immense energy, water, and compute resources that many nations cannot sustain. As models scale and data centres expand, the risk is an AI ecosystem that falters under resource stress and leaves large parts of the world behind. The challenge now is to make resilience a core design principle: building energy-e[icient models, frugal and modular architectures, and green, locally adaptable infrastructure. This panel will explore how countries can embed resilience across the AI stack and ensure that innovation remains sustainable, reliable, and accessible in a resource-constrained world.

• How do we create market incentives for companies to favour efficient, distilled, domain-specific models over brute-force scaling?
• Can modular, edge-first architectures break the dependence on mega data centres and shift innovation power back to the Global South?
• Can nations align on a shared vision of energy-aware AI, or will industrial policy competition derail cooperation?
• Is “frugal-by-design” AI a viable alternative to trillion-parameter models, or will the push for e[iciency inevitably compromise accuracy, reliability, and performance?

Closing Session

Closing Remarks & Vote of Thanks