Nasscom, in collaboration with RBB Economics, convened a focused discussion on "From Framework to Action: Implementation of the AI Chakras" in New Delhi on April 15, 2026.

The discussion was built on the momentum of the AI Impact Summit 2026 and the New Delhi Declaration. As India moves from articulating principles to enabling implementation, the focus is steadily shifting towards operationalising the "AI Chakras": a framework that captures the key pillars required to build a robust, inclusive, and future-ready AI ecosystem. The objective of this convening was to ground these principles in real-world market dynamics, infrastructure constraints, and policy pathways.

The session saw participation from a diverse set of stakeholders, including representative from technology companies, infrastructure providers, startups, and policy experts. This multi-stakeholder representation ensured that the discussion was both comprehensive and grounded in practical realities across different layers of the AI ecosystem.

The deliberations centred around two critical AI Chakras: Resilient, Efficient & Innovative AI Systems, and Democratising AI Resources

Democratising AI Resources: Enabling Fair and Competitive AI Markets

During the session, discussions on competitive markets highlighted the evolving structure of the AI ecosystem across different layers of the stack. While India is witnessing strong innovation and competition at the application layer, particularly through mobile-first deployments, upstream segments such as foundation models and infrastructure remain relatively concentrated. This has implications for long-term market contestability and access to critical AI resources. The conversation also explored the role of existing competition law frameworks in addressing emerging concerns, alongside the need to ensure that regulatory responses remain proportionate and forward-looking.

EnablingResilient, Efficient & Innovative AI Systems: Infrastructure at Scale

The second set of discussions focused on the infrastructure backbone required to support India's AI ambitions. Panellists emphasised that AI resilience extends beyond technological capabilities and is deeply linked to physical and digital infrastructure systems. Key issues included vulnerabilities in sub-sea cable networks, delays in cross-border repair mechanisms, gaps in power distribution and grid readiness, and capacity constraints in data centres. The conversation also highlighted the increasing complexity of semiconductor supply chains and the need for continuous innovation to support AI workloads.

At the same time, the discussion reflected a pragmatic shift in approach towards AI development. Rather than focusing solely on building large foundational models, there was an emphasis on execution, deployment, and the growing relevance of smaller, context-specific models tailored to diverse use cases. As AI systems become more integrated into mission-critical sectors, the need for reliability, accountability, and workforce readiness also emerged as key priorities.

Key Outcomes

• On the state of competition in the AI market, it was discussed that India's AI ecosystem shows active competition at the application layer but weaker domestic presence at the foundation model layer compared to markets like Japan and Korea. Concentration risks at the infrastructure and foundation model layers warrant close monitoring.
• With respect to existing governance frameworks, the Competition Amendment Act 2023, particularly its provisions on intent to collude, offers a sound basis for addressing AI market concerns. Therefore, the priority should be leveraging existing frameworks effectively rather than introducing new legislation.
• For critical AI resources such as compute, chips, and foundation models, competitive markets may need to be actively supported through targeted policy intervention to ensure broad and equitable access.
• AI resilience is fundamentally an infrastructure challenge, not just a technology issue. Building resilient AI infrastructure requires addressing foundational gaps in power distribution, grid readiness, sub-sea cable networks, and supply chain vulnerabilities in parallel with rising compute demand.
• On AI models, the focus should shift from building standalone foundational models to effective execution and deployment, with increasing relevance of smaller, use-case-specific models based on local and context-based data.