India’s AI Growth Needs More Than Just Metropolitan Hubs
By Amit Kapoor and Mohammad Saad
Large urban centres are leading the way in creating value in India’s AI story. The economic value from AI is rapidly clustering in these metropolitan cities and smaller cities risk getting sidelined. For India, it is crucial that AI development is inclusive and Tier 2 & 3 cities equally benefit from the value that AI generates. Moving forward, the vision for these smaller cities must be grounded in developing indigenous and region-specific AI models that cater to the specific needs of their informal and small-scale businesses. Achieving this will require AI policies that not only differ from those designed for metropolitan innovation hubs but are also aligned with the realities of smaller cities.
When it comes to AI development in metropolitans, the primary constraints are limited to AI compute deficits, data availability, workforce shortage and limited R&D. However, the challenges for smaller cities are quite different and fundamental. Tier 2 and 3 cities struggle with structural barriers including limited internet and broadband connectivity, inadequacy of supporting infrastructure and affordability concerns.
Despite deep internet penetration in India, the quality of the internet remains uneven, with significant regional disparities. According to the Ookla Speed test Global Index, Mumbai ranked 123rd out of 200 cities for fixed broadband speeds in February 2025. Mumbai recorded download speeds of 58.24 Mbps. If a large metropolitan area faces such speed constraints, the challenges are likely to be severe in smaller towns. The penetration of high-speed fibre connection is also low for urban centres. Research from Ashoka University shows that in urban areas, which include Tier 2 and Tier 3 cities, only 15.3% of households have access to high-speed fibre connections. Low high-speed fibre penetration and poor internet quality can have a direct impact on AI development prospects.
For a technology such as AI, which requires substantial investment in compute infrastructure, smaller cities have particularly low compute capacity. Colliers’ report – “The Digital Backbone: Data Centre Growth Prospects in India”, found that these cities account for only 6% of India’s total data centre capacity, approximately 82 MW. Limited local compute capacity constrains data processing and increases latency for cloud-based services.
A lack of digital and higher-order skills exacerbates AI adoption challenges further. Research from Ashoka University indicates that the most common barrier to internet adoption is low digital readiness. One in two rural households and two in five urban households are without internet access. These households report that they either do not know how to use the internet or are unaware of its potential uses. Since 40% of urban households lack basic internet usage skills, it is likely that this share would be higher in smaller urban centres.
Cost barriers represent another major obstacle in AI adoption for small-scale informal businesses that dominate Tier 2 and Tier 3 cities. India has witnessed a steady increase in the number of such firms. According to the Annual Survey of Unincorporated Sector Enterprises, the number of informal enterprises rose by 9%, from 59.7 million in 2020–21 to 65 million in 2022–23. A NITI Aayog report found that 59% of India’s small-scale businesses face financial constraints that limit their ability to invest in AI, including the high costs of AI tools, computer infrastructure, and training. The report also notes that 91% of MSMEs believe that AI should be democratically available and affordable.
Given these constraints, developing AI ecosystems in smaller cities requires a phased approach. In the short run, these cities must focus on improving connectivity, digital skills and access to affordable AI tools. Over time, they can build the foundations for local AI innovation.
At present, progress on sovereign AI at the national level remains limited. However, India has made notable advances in AI applications and wrapper-based innovations. In the short to medium term, if policymakers manage to provide internet access, skilling, alongside gradual expansion of compute infrastructure, smaller cities can serve as absorption points for better-funded AI startups based in metropolitan areas. These startups can adapt applied AI solutions and wrappers to local contexts, enabling AI penetration. However, it is critical that low-cost incentives are provided to emerging businesses in metropolitan cities so that setup costs in smaller cities are manageable and adoption barriers are reduced.
In the long run, the goal for Tier 2 & 3 cities should clearly shift towards fostering locally rooted startups that develop localized AI systems. Firms from metropolitan areas that initially provided solutions to smaller towns would have already generated sufficient skill exposure among local workers to enable the formation of original enterprises. Additionally, growing compute infrastructure would provide a foundation for more advanced AI development. However, substantial support for early-stage startups will remain crucial.
At present, the AI startup environment in India remains relatively unfriendly to innovation. Early-stage AI startups face heightened risks due to high compute costs, increasing technical complexity, and limited resources for continuous product iteration, often resulting in capital shortfalls. As a result, investors are often hesitant to fund experimentation and research-heavy AI startups. Government funding for critical early-stage startups is therefore necessary to sustain and nurture them during this formative phase.
Ultimately, India’s AI growth cannot remain concentrated only in metropolitan centres. If smaller cities continue to face gaps in connectivity, infrastructure, skills, and finance, they risk being left behind in the country’s digital transformation. In the short term, improving internet access, digital skills, and affordable AI adoption can help these cities absorb innovations developed in metros and bring practical AI tools to local businesses. Over time, the focus should shift towards nurturing local startups that build region specific AI solutions, ensuring that Tier 2 and Tier 3 cities are meaningfully integrated into India’s AI ecosystem and share in the economic value it creates.
The article was published with Financial Express on March 7, 2026.
