STRATEGIES AND BARRIERS FOR GREEN AI ADOPTION: A SURVEY OF ENERGY-EFFICIENT DEEP LEARNING PRACTICES IN PAKISTAN'S TECH ECOSYSTEM

Abstract

The emergence of artificial intelligence (AI) has become an issue of concern because of its effect on the environment, especially with the consumption of a lot of energy in deep learning models. This paper explores the strategies and obstacles that affect the use of Green AI practices by AI professionals in the technology ecosystem in Pakistan. The survey was performed using a quantitative method in the survey of 100 software house and start up and research institutions professionals. The analysis of data was done through PLS-SEM that would facilitate the research of the relations between Sustainability Awareness, Technological Readiness, Financial Constraints, Organizational Sustainability Strategies, and Green AI Adoption. The results show that Sustainability Awareness, Technological Readiness, and Organizational Sustainability Strategies have a positive impact on adoption, whereas Financial Constraints have negative implications on implementation. The paper will offer useful suggestions to organizations, policymakers, and stakeholders to improve energy-efficient AI activities and also add to the theoretical literature on sustainable AI implementation in developing economies.