India's Strategic AI Roadmap: A Phased Approach to Avoid Premature Lock-ins
The Economic Survey released on 29 January 2026 has outlined a carefully sequenced strategy for artificial intelligence (AI) development in India, emphasizing the need to avoid premature regulatory overreach or technological lock-ins. The nation's approach focuses on building coordination mechanisms first, developing capacity next, and implementing binding policy leverage last, allowing institutions and markets to co-evolve organically.
Phase One: Coordination and Experimental Foundation
The initial phase of India's AI strategy centers on operationalizing already announced institutions and aligning incentives to enable widespread experimentation. Policy should facilitate bottom-up innovation by expanding the reach of existing shared infrastructure under the ambitious IndiaAI Mission.
This foundational approach includes several key components:
- A government-hosted, community-curated code repository for collaborative development
- Pooled access to public datasets through structured initiatives
- Shared access to computing infrastructure already underway
- Clear focus on application- or sector-specific, small and open-weight models for efficient resource utilization
Medium-Term: Selective Scaling and Regulatory Framework
Once coordination mechanisms become functional and early experimentation generates sufficient evidence, policy can shift toward selective scaling in the medium-term. This phase envisions expanded shared and certified domestic computing infrastructure with voluntary participation by large, resourceful firms linked to regulatory facilitation and access to public datasets.
Simultaneously, AI regulation should be formalized on a risk-based and proportionate basis, featuring:
- Graduated obligations for AI firms codified according to scale and sector of use
- Oversight embedded within existing sectoral regulators rather than through a single omnibus AI law
- Deepening of the AI Safety Institute's role from analysis to structured scenario testing, red-teaming, and international cooperation
- Clearly articulated non-negotiable boundaries for high-risk applications
Long-Term Goals: Resilience and Adaptation
India's long-term AI objectives encompass two primary areas of focus. First, the nation must shift toward building resilience, particularly regarding access to advanced computing hardware through strategic partnerships and diplomatic efforts. The objective is to reduce India's vulnerability to external shocks in the global technology landscape.
Second, sustained adaptation of labour markets and education systems will be essential. Primary education must prioritize foundational cognitive and socio-emotional skills, while skilling systems must align themselves with both AI- and human-centric sectoral requirements.
Strategic Considerations and Global Context
Given constraints related to capital, computing capacity, energy, and infrastructure, pursuing scale for its own sake is neither efficient nor necessary for India. Instead, the chapter makes the case that a bottom-up, multiple sector-specific approach under a single vision has the potential to pay significant dividends and create dignified employment opportunities for India's youth.
India's development of AI must be grounded in open and interoperable systems to promote collaboration and shared innovation. This pathway aligns more closely with India's strengths in human capital, data diversity, and institutional coordination.
Data Governance Framework
The proposed framework for data governance strikes a careful balance between openness to cross-border flows and strengthening accountability and regulatory visibility. It is rooted in the objective of ensuring that the value accruing from India's domestic data is retained within the country for the benefit of its people.
The government's role is framed as that of an enabler and coordinator, helping markets and institutions adjust in step with technological change. Overall, the chapter treats AI as a strategic choice, with the central message being that India's opportunity lies in deploying AI in a way that is economically grounded and socially responsive.
Global AI Adoption Trends
AI is no longer a distant or speculative technology. It is increasingly being adopted, even if in an experimental capacity, in organizations around the world. Based on a survey of 1,993 firms by McKinsey, 88% of organizations surveyed in 2025 reported that they are utilizing AI in at least one of their business functions.
Of those using AI:
- 31% are in the process of scaling its application across the organization
- 7% have already fully deployed and integrated AI
Innovations and continuous improvement in AI capabilities are driving firms and new startups to develop ways in which AI can be applied to solve real-world problems. At the same time, greater visibility into AI adoption has brought greater clarity on the nature of the technology itself.
Concentration and Labour Market Considerations
Over the past year, it has become evident that while the use of AI tools can be widespread, the frontier of AI remains highly concentrated. The development and training of advanced foundational models is increasingly capital-, compute-, data- and energy-intensive, favoring a small set of firms with access and the political capital to secure large-scale infrastructure projects, specialized hardware, and deep pools of technical talent.
Early evidence has also begun to temper some of the more extreme predictions surrounding AI's near-term labour impact. For instance, a study conducted by Yale's Budget Lab indicates that the broader labour market in the United States has not experienced a discernible disruption due to AI.
This does not invite complacency, especially from a policymaker's perspective. While labour may be complemented in the near term as organizations work to incorporate AI into their tasks, productivity gains from augmentation have a ceiling. All in all, caution is still warranted as India attempts to solve the puzzle of AI and labour, which represents one of the most considerable looming uncertainties about the technology.
