Exploring the sustainable scaling of AI dilemma: a projective study of corporations’ AI environmental impacts

Abstract

The rapid growth of artificial intelligence (AI), particularly Large Language Models (LLMs), raises environmental concerns extending beyond greenhouse gas emissions. Limited transparency from major providers hinders companies’ ability to evaluate their AI-related environmental impacts and achieve their net-zero targets. This paper proposes a methodology to estimate the environmental impact of a company’s AI portfolio, delivering actionable insights without requiring deep AI or Life-Cycle Assessment (LCA) expertise. Using a modeling approach that incorporates increased AI usage, hardware efficiency, and electricity mix changes aligned with IPCC (Intergovernmental Panel on Climate Change) scenarios, we forecast AI impacts through 2030. Under a high adoption scenario, driven by complex generative AI models and frameworks, AI electricity consumption could rise by a factor of 24.4. Mitigating this impact by 2030 demands coordinated efforts across the AI value chain. Isolated measures like hardware or model efficiency alone are insufficient. We call for standardized environmental assessment frameworks, increased transparency, and a “Return on Environment” metric to align AI development with sustainability goals.