Here’s why India’s AI dividend will be earned by legacy manufacturers rather than digital-natives

But even as the summit calls for a more democratic and socially relevant AI, the overarching question is this: Where will AI actually move the needle?

Will it be confined to elite technology ecosystems, empowering those already at the top of the pyramid? Or will its true dividend come from transforming our vast education, health and agriculture ecosystems—or legacy manufacturing sectors that employ millions?

AI is already delivering value across sectors—improving productivity, democratizing access and aiding livelihoods. In healthcare, AI-assisted diagnostics improve early detection and clinical decision-making.

In agriculture, precision advisories and crop intelligence tools help farmers increase incomes and manage risk.

In education, adaptive systems personalize learning at scale. Even ‘small AI’—narrowly designed, task-specific systems—is demonstrating measurable impact, reducing delays, errors and manual workload in daily operations.

These are important gains. But India’s long-term economic shift cannot be determined solely by digital-native businesses.

It will eventually be shaped by whether AI can modernize legacy industries like textiles, food processing, leather, auto components, light engineering and chemicals—sectors where depth of capacity exists, but production processes are often uneven, partly digitized or manually monitored.

From cost competitiveness to capability competitiveness: India’s manufacturing base is dominated by micro, small and medium enterprises (MSMEs). These industries employ millions, support small industrial networks and drive exports.

Yet, many operate with informal workflows and thin margins, competing on low labour expenses and homegrown adaptations to cut costs. Productivity improvements here matter far more for national income than incremental gains in already digitized sectors.

This is not about replacing labour at scale. The opportunity is different—embedding intelligence into existing systems to reduce defects, limit material waste, lower energy consumption and strengthen audit preparedness.

Take textiles. AI-enabled quality inspection can detect weaving or dyeing defects in real time.

Predictive maintenance tools can reduce machine downtime. Real-time monitoring of water, chemical and energy use can shift sustainability from periodic reporting to measurable reductions in input costs.

Each gain may appear incremental at the unit level. But across thousands of MSMEs, the cumulative effect on output quality, operating margins and export reliability can be significant.

This is where the economic logic becomes clear. Global competitiveness increasingly depends not only on scale or wage arbitrage, but on consistent quality, delivery reliability and transparent supply chains.

International buyers now demand granular visibility across production networks. AI-driven measurement systems allow Indian firms to meet these standards at lower compliance and verification costs.

The sustainability dimension is equally important. As global carbon regulations tighten, data-backed traceability will become a market-entry requirement.

Real-time measurement systems can help firms track emissions, water use and energy intensity continuously rather than assemble documentation retrospectively.

When seen in this context, AI is not just a technology upgrade. It is a competitiveness strategy. However, for AI to succeed in legacy sectors, five implementation lessons are critical.

First, build digital foundations before deploying AI: Readiness will determine what scales. Most MSMEs lack structured reliable data. Without basic digitization—sensors, digital production logs, ERP-lite systems—AI models have little meaningful input.

Therefore, transformation must begin with narrow-scope, low-investment interventions that resolve immediate bottlenecks and accommodate initial resistance.

Second, ensure a visible and time-bound return on investment: MSMEs operate on tight working-capital cycles. AI solutions must deliver measurable reductions in defect rates, energy bills or downtime within months. Long-horizon transformation narratives will not drive adoption.

Third, adopt cluster-level deployment: India’s manufacturing strength lies in clusters—from Tiruppur to Surat to Ludhiana. Shared digital infrastructure, common data standards and collective service platforms can lower adoption cost and create scale efficiencies.

Fourth, prevent vendor lock-in: MSMEs must retain ownership of their operational data. Open standards and interoperability are essential to avoid long-term dependence on proprietary systems that may increase costs.

Fifth, use AI as a productivity-enhancing tool, not labour-displacing: The focus needs to shift from headcount management to productivity and value upgradation. Adoption improves when AI improves output quality without impacting jobs.

These lessons have direct implications for industrial policy. Incentives should reward measurable productivity improvements rather than mere technology acquisition.

Credit-linked support can accelerate basic digitization. Industrial parks and manufacturing clusters should embed digital infrastructure at the design stage instead of retrofitting later.

Public policy can enable cluster-wide digital ecosystems rather than subsidize isolated pilots. Markets should function efficiently to ensure capital and technology access to smaller firms.

If AI can reduce defects across thousands of factories, reduce energy intensity at scale and shorten compliance cycles, the cumulative effect on margins and exports could be substantial.

The next wave of AI growth will not come only from code. It will come from machines, shop floors and supply chains becoming intelligent.

India’s AI moment will be won—or lost—on the factory floor.

These are the author’s personal views.

The author is additional secretary, ministry of textiles.