High Profile Area at the

ComAI Lecture: "Between Power and Nature: Provocative reflections for an Eco-Political Economy of AI" (Prof. Dr. Benedetta Brevini)

Abstract: Despite the growing concern over the environmental harms of ICT systems (Ferreboeuf, 2019) Artificial Intelligence (AI) gets principally heralded as the key technology to solve contemporary challenges, including the Climate crisis, which is one of the goals of sustainable development (Saetra, 2022). This AI hype frequently overlooks the significant environmental impacts stemming from the increasing demand for AI tools (Brevini, 2020a; Dobbe, 2022). Amid the media frenzy promoting the popularity of AI models, major Digital Lords (Brevini, 2020b) like Microsoft, OpenAI, and Google on the one hand and various international institutions (OECD, 2022; European Commission, 2022) have finally acknowledged the substantial environmental costs associated with meeting the increasing demand for AI tools. These costs include unparalleled demand of rare metals (European Commission, 2022), massive energy expenditure (Brodie,2023) and an unprecedented impact on water consumption (George et al., 2023). Microsoft's most recent environmental report for 2022 after the launch of Open AI generative AI services reveals a significant 34% increase in its worldwide water consumption from 2021 to 2022, reaching nearly 1.7 billion gallons (Microsoft, 2022).

Surprisingly, despite a wealth of evidence (Brevini, 2021), there is no universally agreed-upon conceptual framework or standardized guidelines for understanding the intricate ecological impacts of AI. The widespread adoption and perceived benefits of AI, along with the distinct separation between research fields, contribute to a fragmented academic landscape.

In various interventions, I called for the development of what I named "an Eco political economy of AI" (Brevini, 2021; 2022; 2024) to understand the complex elements involved in the assessment of AI environmental harms. This approach involves examining three crucial segments of the extractive global production and supply chain of AI to account for its environmental costs: Mining and resource extraction; b) Consumption, energy use and carbon footprints; c) Digital waste. In this lecture, I delve deeper into my ideas, presenting a nuanced elaboration of this framework. By integrating theories from Media and Communication, geography, computing, and engineering with indigenous concepts and environmental justice paradigms, the Eco Political Economy of AI offers the tools necessary to position the Climate Crisis at the forefront of technology and AI development.

Biographical note: Benedetta Brevini is Visiting Professor at New York University, Institute of Public Knowledge and Associate Professor of Political Economy of Communication at the University of Sydney. Before joining the academy, she worked as journalist in Milan, New York and London for CNBC, RAI and the Guardian. She is the author of several books including Is AI good for the Planet (2022), Amazon: Understanding a Global Communication Giant (2020), Public Service Broadcasting online (2013) and the editor of Beyond Wikileaks (2013), Carbon Capitalism and Communication: Confronting Climate Crisis (2017), Climate Change and the Media (2018). She is currently working on a new volume for Polity entitled "Communication systems, Technology and the climate emergency".