India won’t be judged by how fast it builds frontier AI, but by how quickly it makes AI usable, affordable, and routine for its smallest firms
India’s micro, small and medium enterprises (MSMEs) sit at the heart of its competitiveness story, but the next leap will come from democratising AI so that advanced capabilities become usable, affordable, and routine for the smallest firms. Official estimates underline this centrality: MSMEs account for about 30.1 percent of GDP and 45.73 percent of exports, while India’s formalisation drive has expanded the recorded base to roughly 65 million MSME units. In that sense, the real question is no longer whether MSMEs should adopt AI, but whether India can democratise AI so that sophisticated capabilities become usable and affordable for the smallest firms, in ways that compound productivity while advancing inclusion.
The real question is no longer whether MSMEs should adopt AI, but whether India can democratise AI so that sophisticated capabilities become usable and affordable for the smallest firms, in ways that compound productivity while advancing inclusion.
This framing is strongly reinforced by India’s plan to convene the AI Impact Summit from 16-20 February 2026 in New Delhi, which is built around the theme of democratising access to AI and achieving measurable impact across sustainable development objectives. The implication is straightforward: if AI leadership is defined only by frontier models or elite enterprise deployments, it will remain disconnected from development priorities. If it is defined by diffusion at the productive edge of the economy, then MSMEs become the decisive proof-point. The credibility test, therefore, is whether AI can improve the everyday decisions that determine survival and scaling for small firms through compliance, quality, inventory, hiring, energy use, supplier reliability, and buyer discovery.
For MSMEs, AI is best understood as an operational agenda that lowers the thresholds of expertise, time, and cost required to access reliable decision support. For most MSMEs, the most realistic pathway is through AI embedded into workflows that owners and supervisors already run, rather than AI as a transformation programme requiring new teams, high fixed costs, or heavy integration. This is precisely why adoption is likely to be shaped by plug-and-play tools: AI-enabled customer support, assisted marketing and content creation, demand forecasting, inventory optimisation, basic analytics, and computer-vision quality checks that raise reliability while reducing waste. For example, AI-enabled inspection can materially improve defect detection in applied settings, which matters directly for standards compliance and export readiness.
At the same time, the adoption gap is now clearer than the awareness gap. Evidence from a NASSCOM–Meta collaboration reports strong belief among tech-enabled MSMEs that AI can drive growth and improve productivity, but it also highlights practical frictions that block diffusion — notably limited awareness of the right tools, affordability constraints, and budget limits that deter sustained use. This is important analytically because it shifts the policy challenge away from persuasion and toward market-shaping: lowering first-mover risk, improving discoverability of trustworthy tools, and ensuring that early experimentation does not become a cost trap. Complementing this, estimates suggest AI can raise MSME productivity and reduce operating costs through process optimisation and data-driven decisions, but these gains depend on whether adoption becomes routine rather than exceptional.
Figure 1: What India’s Tech-enabled MSMEs Say about AI (selected survey findings)
Source: Author’s own, data from NASSCOM–Meta White Paper
The more fundamental point is that democratisation is not primarily about persuasion, but about building a diffusion pathway that makes AI repeatable for firms that are time-poor, capability-constrained, and often compliance-burdened. In such settings, value is created through the compression of transaction costs: fewer filing errors, faster documentation, reduced downtime, improved quality control, better matching of products to buyers, and lower-cost customer engagement. This is also why frugal AI matters as a design principle for India, with application-specific, smaller models and deployments that work within constraints of cost, compute availability, power, and connectivity rather than assuming GPU-heavy stacks and perfect data. The Economic Survey 2025–26 flags “local ingenuity and frugal AI” as an emerging bottom-up pathway in India, and that is precisely the route that can make AI usable for MSMEs as a routine capability rather than a high-risk, high-fixed-cost bet.
The more fundamental point is that democratisation is not primarily about persuasion, but about building a diffusion pathway that makes AI repeatable for firms that are time-poor, capability-constrained, and often compliance-burdened.
India’s AI democratisation agenda is often constrained by factors rooted in India’s enterprise structure, where the median firm is small, resource-constrained, informal or semi-formal, and operates in multilingual markets. The core question is not whether India can build frontier AI, but whether it can make AI usable and routine at the productive edge of the economy, especially for MSMEs.
First, India’s diffusion advantage lies in public digital rails that reduce transaction costs and enable scale, creating conditions for AI adoption as an incremental capability rather than a one-time “big transformation”. The state’s AI-facing initiatives signal this direction, including efforts to broaden access to computing and shared resources through national platforms. Government communication on the IndiaAI Mission emphasises large-scale compute provisioning and shared assets intended to lower entry barriers beyond elite users. For MSMEs, the point is practical where these public goods can shift AI from a high fixed-cost bet into something that can be piloted, evaluated, and scaled with lower risk.
Second, language and local context are not peripheral to democratisation in India; they are the adoption infrastructure. AI diffusion stalls if tools remain English-first and metro-first, because many MSME decisions are made in regional environments where literacy, time, and formal managerial training are unevenly distributed. India’s emphasis on language AI and translation infrastructure, especially the BHASHINI platform under the National Language Translation Mission, creates a practical adoption layer for MSMEs by enabling speech, translation, and Indian-language interfaces through shared public resources that apps and services can integrate.
Language and local context are not peripheral to democratisation in India; they are the adoption infrastructure. AI diffusion stalls if tools remain English-first and metro-first, because many MSME decisions are made in regional environments where literacy, time, and formal managerial training are unevenly distributed.
Third, inclusion is the hinge that links AI diffusion to India’s growth strategy and social outcomes. If AI adoption concentrates in firms that already have capital, English proficiency, and managerial slack, it will widen productivity and market-access gaps. If, instead, AI becomes accessible to microenterprises and women-led firms, it can strengthen livelihoods and job quality, expand formal market participation, improve compliance without becoming more punitive, reduce waste and raise resource efficiency, and widen access to finance and services. As per the Ministry of MSME’s, women-owned units account for 20.5 percent of Udyam-registered MSMEs since 1 July 2020, but account for 18.73 percent of employment and 11.15 percent of investment, which aligns with many women-led enterprises remaining micro in scale and facing layered constraints such as credit gaps, skills constraints, and weaker market linkages.
Figure 2: Women-owned enterprises in India’s MSME ecosystem
Source: Author’s own, data from Ministry of MSME, Udyam Registration Portal & Udyam Assist Platform
Finally, India’s AI Impact Summit should be treated as an invitation to move from “AI potential” to “AI diffusion architecture”, where impact is measured not only by model sophistication but by who benefits, at what cost, and with what productivity and inclusion effects across the enterprises that anchor India’s real economy. The strategic payoff is concrete: when AI becomes an everyday capability for small firms, productivity gains compound, quality improves, compliance becomes less punitive, and market access widens, supporting a pathway to growth that is both more competitive and more socially sustainable.
Soumya Bhowmick is a Fellow and Lead, World Economies and Sustainability at the Centre for New Economic Diplomacy (CNED) at the Observer Research Foundation.
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.
Dr. Soumya Bhowmick is a Fellow and Lead for World Economies and Sustainability at the Centre for New Economic Diplomacy (CNED) at the Observer Research ...
Read More +
PREV