Impact of Geographical Location and Energy Sources on the Carbon Footprint of AI Models- A Survey-Based Study

Abstract

Abstract Artificial Intelligence (Al) has become an indispensable technology driving automation, prediction, and optimization across industries. However, the unprecedented computational demand of training and operating large Al models results in significant electricity consumption and greenhouse-gas emissions. This paper presents a comprehensive analysis of the carbon footprint of Al model development and deployment with particular attention to geographical location and energy-source variation. A quantitative survey of eighty 80 respondents including students, professionals, and Al practitioners was conducted to evaluate awareness, energy usage respondents including students, professionals, and Al practitioners was conducted to evaluate awareness, energy-usage patterns, and sustainable practices. Statistical and visual analyses reveal that although 65% of respondents recognize Al’s environmental impact, only a small portion of organizations monitor emissions or rely on renewable energy.The work proposes a practical emission- estimation formula. summarizes correlations among awareness ,traning durtion, and organizational monitorine, and outlines policy and technical recommendations for “Green AI.” Keywords Artificial Intelligence. Carbon Footprint, Green Al. Enerey Sources, Sustainability, Data Centers