by Abdullah Hasan Safir and Sanjay Sharma
Safir, A. H., & Sharma, S. (2025). Networks, narratives and neocoloniality of AI for Climate Action. Big Data & Society, 12(2), 20539517251334095. https://doi.org/10.1177/20539517251334095. (Original work published 2025)
Our paper problematises AI for Climate Action by challenging the assumptions and normative beliefs around AI’s promise to ‘solve’ global climate problem.
We ask two fundamental questions to investigate geopolitical power asymmetries around this practice: First, who is doing AI for Climate Action? and second, what discourses are they making?
Our findings reveals that tech corporations from the Global North are leading the use of AI for Climate Action, with global non-profits and civil society organisations as their facilitators. They use certain discourses such as ‘leverage’, ‘innovate’, and ‘responsibility’ to justify their actions and capitalistic, profit-driven goals – in disguise of their benevolent efforts to address climate problems.
The Global South countries can face challenges in accessing digital infrastructures for AI innovation for climate action due to high entry costs and Global North-centric corporate interests. In addition, high-level risk associated with AI applications could cause more harm than benefits for them.
Understanding the Global North dominance and their exploitative narratives is important. Many scholars suggest that climate problem is a colonial problem and show that the Global South is disproportionately affected by climate vulnerabilities. Drawing parallel, we frame AI for Climate Action as a neocolonial practice in our work. If AI technologies are not designed in response to Global South contexts and if the control of these applications remains in the Global North, it would further exacerbate global climate injustice.
Our work also offers methodological innovation. It combines a digital method, ‘issue mapping’ (inspired from Actor-network Theory in STS), with Critical Discourse Analysis (CDA). Critical AI scholarship will increasingly require such inter-disciplinary approaches like the one we have applied to address the complex crises induced by AI.
