Everyone is racing to deploy AI for sustainability. I think we're asking the wrong question

The world doesn't have an AI problem. In developing nations, we have a data problem,** **and no amount of sophisticated machine learning fixes a broken foundation.

Here's what the global conversation gets wrong:

AI is genuinely powerful. By end of 2025, roughly 30% of smart city applications are expected to integrate AI, optimizing energy grids, predicting floods, monitoring air quality. In cities across developing nations, ML models are already forecasting pollution levels with remarkable accuracy, and AI pilot programs are mapping sanitation risks in the most vulnerable urban neighbourhoods.

But zoom out. Urban datasets, especially in developing nations, routinely suffer from incompleteness, inconsistency, fragmentation, and temporal gaps that directly compromise the reliability of any AI model built on top of them.

You cannot build a smart city on a data desert.

The on-ground reality of developing nations

We celebrate smart city missions while ground-level urban data, on informal settlements, waste streams, actual energy consumption in low-income areas, is either absent, outdated, or siloed across multiple government departments that don't talk to each other.

Across low-developed nations, years after launching national AI strategies, critical gaps persist across data infrastructure, AI workforce skills, and ethical and legal frameworks, with no regulatory or ethics framework yet formed. High willingness to adopt AI exists across sectors. The data and systems to feed it responsibly? Still catching up.

For key SDGs like Climate Action, Gender Equality, and Institutions, less than half of 193 countries have internationally comparable data since 2015, and less than 30% of the latest SDG data comes from 2022 or 2023. How do we build AI systems to solve problems we can't even measure?

And here's the irony nobody talks about

A review of 792 studies on AI in sustainability found that only a few actually combine advanced AI applications with deep sustainability expertise. We have technologists who don't understand urban systems, and urban planners who don't trust the tech. The gap is not just in data, it's in disciplines.

So what do I actually believe?

As someone who sits at the intersection of policy, technology, and urban development, I believe the most urgent investment in developing nations isn't the next AI model. It's the boring, unglamorous work of building open data ecosystems, standardising collection protocols, and training local professionals who can interrogate what the AI is actually telling them.

AI is ready. Our data isn't. Our governance isn't. Our institutions aren't.

The nations that will win the sustainability race won't be the ones that deploy AI fastest. They'll be the ones that built the most honest, high-quality, community-rooted data infrastructure first.

That's the work. That's what I'm focused on.