Cooling cities, not just algorithms: Making AI work for India’s most vulnerable
India stands at a turning point. Its AI-powered urban future will be judged not by speed or scale, but by whether it reduces vulnerability.
[...]
Artificial intelligence (AI) is rapidly shaping how cities plan and deliver services. The key question is not whether AI will drive India’s climate response, but whether it can do so equitably—protecting women, informal workers, and the urban poor.
[...]
Delhi’s 2025 Heat Action Plan (HAP) represents a new generation—embedding AI and satellite data to create building-level vulnerability maps. Collaborations with IIT Mandi and partners like Resilience AI help identify the hottest clusters, enabling neighborhoods such as Vivekananda Camp to install low-cost measures—reflective paints, shade structures, and drinking water points.
[...]
A philanthropy representative warned against the illusion of completeness in digital dashboards. Clean maps may hide unregistered slums or waste piles on unmapped streets. “AI cannot replace human judgment,” she argued; inclusive design requires community voices. From the tech side, a geospatial AI manager echoed the “garbage in, garbage out” warning and raised another concern: AI’s own energy footprint. The computing power behind massive AI models consumes energy comparable to parts of large cities—so resilience planning must also limit AI’s environmental costs.
[...]
AI tools can also directly reduce inequality. Equity analytics identify zones where high heat risk coincides with poor access to cooling or healthcare, guiding municipal investments in trees, water, and public cooling centers.
[...]
Still, AI can deepen exclusion if the most vulnerable remain invisible in data. Informal workers, internal migrants, and slum residents often do not appear in official records, and women’s domestic or informal work rarely features in occupational data. Without correcting these gaps, AI systems will perpetuate inequality even as they model “solutions.”
The path forward rests on three principles:
- Accurate, context-aware data—ground surveys must complement satellites and need capacity and community partnerships to interpret outputs.
- Collaboration—across public, private, and civic actors—with open, safe data-sharing standards.
- Community co-design and co-creation—those enduring the heat should define what problems AI solves and what interventions fit their realities.
[...]