The Ministry of Electronics and IT's recent constitution of the AI Governance and Economic Group, or AIGEG, is an acknowledgment that artificial intelligence can no longer be treated as a narrow technology subject within government. Questions around AI are now tied to economic growth, labour markets, public administration, national competitiveness and strategic influence at the same time.
The harder part, however, begins after the committees are formed and the initial momentum around summit declarations begins to settle.
India enters the AI transition with real strengths. It has built population-scale digital infrastructure, created public platforms that altered the economics of payments and identity, and developed technical capacity that gives it credibility in global technology discussions. But AI introduces a different set of dependencies.
Compute infrastructure remains concentrated among a relatively small number of global players. Advanced semiconductor ecosystems are still limited to a few geographies, while foundational AI models are increasingly being developed by a small set of large companies. Countries that steadily build domestic capabilities will likely be better placed to capture the broader economic and developmental opportunities emerging around AI.
That is the larger context in which AIGEG's work should be viewed.
The first requirement is to broaden India's AI adoption base across the wider economy. Much of the current conversation around AI still focuses heavily on model development and compute supply, but long-term competitiveness will depend equally on the scale of domestic demand creation.
The more important question is whether Indian enterprises, public systems, MSMEs, research institutions and State departments are able to meaningfully integrate AI into everyday operations and decision-making. That question is closely tied to compute planning itself.
Without a clearer understanding of where AI adoption is likely to deepen across sectors, infrastructure expansion risks becoming disconnected from actual demand. AIGEG could therefore commission a nationwide sectoral study mapping where AI adoption is already generating value, where barriers persist, and which sectors are likely to drive future compute demand at scale.
The second priority is labour transition. Much of the global debate on AI and employment still draws heavily from advanced economies with formal labour protections and mature retraining systems. India's employment landscape is structurally different. Informal work, uneven digital access and regional disparities complicate any straightforward reading of automation forecasts. AIGEG's terms of reference already recognize this by calling for assessment of labour-market impacts across informality, skill diversity and regional variation. The country now needs its own evidence base rather than imported assumptions.
An integrated national survey on AI Exposure and Augmentation would be a useful beginning. At present, workforce datasets, productivity indicators and informal labor registries exist in separate systems. Bringing them together would allow policymakers to identify where AI is likely to displace work, where it may augment productivity and where new employment categories may emerge. It would also help convert an economic mandate into a primary data instrument.
The third area requiring attention is federal coordination. State governments are already deploying AI systems across welfare administration, agriculture, policing and education. The present Group, however, does not include state representation. The omission is worth revisiting. Many implementation failures in India emerge not from weak policy intent, but from uneven coordination between central frameworks and state execution capacity. A standing consultative mechanism with state IT departments and sectoral administrators would make national governance efforts considerably more credible. The success of UPI and the broader DPI story has shown how much smoother implementation becomes when the states are partners from the start.
The fourth suggestion is to make "Defer" credible and time-bound. The Terms of Reference allow AIGEG to classify AI use cases as "deploy", "pilot" or "defer". The first two will run smoothly. The third is politically harder, particularly when investment commitments above two hundred billion dollars have been announced and the New Delhi Declaration is in circulation. Without procedural discipline, "defer" will quietly become "undecided". Every defer ruling should carry three conditions: public notification with reasoning, a sunset of 12-18 months, and a published list of gating criteria that, when met, trigger automatic reclassification. The criteria might include availability of evaluation methods, presence of redressal capacity in the deployment context, or sectoral regulator readiness.
Also importantly, settle a common vocabulary while sectoral work is still in flight. The Reserve Bank has issued its Framework for Responsible and Ethical Enablement of AI. CERT-In has published the AI Bill of Materials. SEBI, IRDAI, and TRAI are each examining AI questions in their respective domains. Each effort is substantial in its own right. The definitions of foundational terms, including high-risk system, auditability, explainability, and incident, are still in formation. A Common Lexicon and Risk Taxonomy, issued early by AIGEG and binding within the Central Government, would help align the sectoral work as it matures, and would save the system the cost of reconciling divergent definitions later.
Finally, India should approach global AI governance with greater strategic clarity. Much of the international conversation is presently shaped by the West. India occupies a different position. It combines democratic politics, linguistic diversity, population-scale digital infrastructure and developmental complexity in ways few countries do. That gives it standing in debates around public-interest deployment, workforce transitions and inclusive AI systems. We should be able to leverage that advantage.
The article has been authored by Srinath Sridharan and Indranuj Pathak, who are corporate adviser and author of 'Family and Dhanda'; and Tech Policy professional respectively.
Disclaimer: The views expressed in this article are solely those of the author and do not necessarily reflect the opinion of NDTV Profit or its affiliates. Readers are advised to conduct their own research or consult a qualified professional before making any investment or business decisions. NDTV Profit does not guarantee the accuracy, completeness, or reliability of the information presented in this article.
ALSO READ: How India's Vegetarian Food-Economy Lost Regional Imagination
Essential Business Intelligence, Continuous LIVE TV, Sharp Market Insights, Practical Personal Finance Advice and Latest Stories — On NDTV Profit.