India's urban AI journey can be a role model for Global South

India is at a turning point in its urban growth and transformation. As projected by the National Commission on Population in 2020, nearly 39.15 percent of Indians are expected to live in cities by 2036. This scale of urbanisation places sustained pressure on transport systems, public safety, sanitation, utilities, and city management. Cities already play a critical role in India’s economic output, with major metropolitan regions such as Mumbai, Delhi, Bengaluru, Chennai, and Hyderabad contributing significantly to national GDP.

Meeting these challenges is not only a matter of capital investment. While World Bank estimates indicate that India will need to invest over $840 billion in urban infrastructure over the next 15 years, outcomes depend equally on how cities manage assets, infrastructure and decision-making processes. In this context, artificial intelligence is increasingly being integrated into urban governance as an operational capability that strengthens responsiveness, efficiency and resilience.

Over the past decade, India’s cities have progressed from basic digitisation towards data-driven and AI-enabled urban operations. National initiatives such as the Smart Cities Mission, AMRUT (Atal Mission for Rejuvenation and Urban Transformation) and the National Urban Digital Mission have created the enabling digital and institutional foundations for this transition. Importantly, AI adoption in Indian cities has focused on embedding intelligence into daily civic functions rather than treating AI as a standalone technology intervention.

In Gorakhpur (Uttar Pradesh), recurrent urban flooding during the monsoon posed a persistent governance challenge due to the city’s low-lying, saucer-shaped geography. Until recently, flood management was largely reactive, with pumps and personnel mobilised only after flooding had already occurred, resulting in response times of 10 to 12 hours. In 2022, the city implemented an AI-enabled urban flood management system that integrated rainfall forecasting, predictive analytics, sensor data and historical flood patterns. Vulnerable wards could be identified up to 24 hours in advance, with automated pump operations supporting preparedness. Emergency response times were reduced to under two hours, pump downtime declined significantly, and authorities were better equipped to manage flooding proactively.

In Prayagraj (Uttar Pradesh), large-scale crowd management posed a distinct challenge. During major religious congregations, the city hosts hundreds of millions of visitors over short periods. In preparation for recent events, including deployments beginning in 2023 and scaled further in 2024, authorities adopted AI-enabled crowd analytics and decision-support systems. Trained on historical footage and local operational inputs, these systems enabled real-time crowd density estimation, prediction of congestion zones and dynamic redirection of crowds and traffic. AI functioned as an operational support layer for administrators, enhancing situational awareness and response capability across a vast geographic area.

In Ahmedabad (Gujarat), traffic congestion and violations put sustained pressure on enforcement capacity. Beginning around 2021, the city deployed an AI-enabled traffic monitoring and enforcement system using computer vision and automatic number plate recognition, integrated with national databases. Implemented across more than 90 intersections using thousands of cameras, the system supported consistent enforcement and contributed to a gradual reduction in violations. Over time, authorities observed behavioural change alongside improved traffic flow, without requiring proportional increase in manpower.

Pimpri-Chinchwad (Maharashtra) illustrates the value of phased AI adoption. Facing public safety and surveillance coverage gaps, the city began deploying AI-enabled systems in 2022, starting with pilot implementations before scaling. Over time, coverage expanded to hundreds of locations supported by fixed cameras, ANPR systems and AI analytics licences.

This incremental approach reduced operational risk, built institutional trust and ensured that AI systems became embedded in routine city operations. Such models are particularly relevant for cities operating under fiscal and administrative constraints.

As IndiaAI continues to shape the national AI narrative, a clear shift is underway from technology deployment towards measurable public impact. As articulated by the leadership of IndiaAI, the focus is on ensuring that AI systems strengthen governance outcomes, improve service delivery and remain grounded in public purpose. AI initiatives are increasingly viewed as part of core urban infrastructure rather than isolated projects, with emphasis on scalability, interoperability and long-term sustainability.

At the same time, data is recognised as the foundational layer for impactful AI. Cities that prioritise data quality, governance and integration are better positioned to deploy AI responsibly and effectively. Leveraging existing digital public infrastructure reduces costs while accelerating adoption.

Equally important is the recognition that AI systems are designed to support public authorities, not replace them. Human oversight, accountability and contextual decision-making remain central to urban governance.

India’s urban AI journey continues to evolve through enabling frameworks, operational learning and continuous refinement, culminating in the India AI Impact Summit (February 16-20, 2026). Rather than offering a single template, India’s experience demonstrates a set of practical pathways for deploying AI in complex, high-density and resource-constrained environments.

For cities across the Global South navigating rapid urbanisation and growing governance challenges, this approach highlights how AI can be embedded into urban systems to enhance resilience, efficiency and service delivery while remaining firmly rooted in local context and institutional realities.

(Sushil Pal, Joint Secretary, Ministry of Electronics and Information Technology, and Aarti Harbhajanka, Co-founder and MD, Primus Partners)