‘The pitch is simple: Generative AI equals efficiencies in time and cost. Professional and educational sectors are widely endorsing and embracing Generative AI. Legal responses are mobilising but are behind the development curve. Most focus so far has been on data rights, IP, biases and possible threats to human agency. Discussions of moral and ethical issues arising from the development of Generative AI have tended to be framed by these concerns in the present. This article argues that equitable downstream cost considerations should be fundamental to Generative AI risk assessment. Through the analysis of the power usage of high-density data centres providing AI services, it is argued that risks to environmental justice presented by AI are real and tangible. Furthermore, it is proposed that governance approaches should account for these as part of environmental impact assessment processes.’
Link: https://www.tandfonline.com/doi/full/10.1080/13600834.2026.2721019
