Generative AIs and the Hidden Cost of Intelligence: A Multidisciplinary Review of LLMs Progress, Resource Consumption and Path to a Sustainable Future

The rapid advancement of Large Language Models (LLMs) and Generative Artificial Intelligence (GenAI) has reshaped technological, business and social environments. Though the models — GPT-4, GPT-5, Claude and others — have shown unparalleled abilities in automation, creative work and decision-making, they have placed enormous burdens on natural resources such as energy, water and computational infrastructure raw materials. This paper offers a multidisciplinary analysis with integrated views from environmental science, computer engineering, economics and business information systems perspective. The research analyzes the environmental externalities of LLMs training and deployment, and measures resource usage patterns and places them in context of global sustainability objectives. Methodologically, the study utilizes a mixed review framework by incorporating systematic literature review and secondary data analysis from peer-reviewed journals in ScienceDirect, Taylor & Francis, Wiley Online, IEEE, MDPI, SpringerLink, SAGE, Emerald and other trustworthy databases. The findings show that the growth of GenAI is happening at an unsustainable level unless checked by strategic interventions on infrastructure efficiency, policy regulation and corporate environment responsibility. It is suggested to imperatively reconcile AI innovation with the circular economy and carbon-neutral computing principles to guarantee that intelligence augmentation is not at the cost of planetary well-being.

 

Keywords: Generative AI, Large Language Models, artificial intelligence, sustainability, energy consumption, energy footprint, carbon footprint, carbon emission, water consumption, multidisciplinary review