Responsible AI deployment in sustainable project execution: Ensuring transparency, carbon efficiency and regulatory alignment

Tettey, Diameh Jacob and Chrisben, Daniels and Victoria, Mbagwu Chinenyenwa and Chukwubuikem, Edike Precious and Abdullahi, Abdulsalam (2025) Responsible AI deployment in sustainable project execution: Ensuring transparency, carbon efficiency and regulatory alignment. International Journal of Science and Research Archive, 14 (3). pp. 1686-1705. ISSN 2582-8185

The convergence of artificial intelligence (AI) and sustainability in project execution presents both transformative potential and ethical complexities. As AI systems are increasingly embedded into infrastructure, construction, energy, and environmental management projects, ensuring responsible deployment becomes imperative to safeguard against unintended consequences, greenwashing, and regulatory breaches. In particular, the triple imperative of transparency, carbon efficiency, and regulatory alignment now defines the benchmark for ethical and sustainable AI integration. This article examines the role of responsible AI in driving sustainable outcomes across project lifecycles, from design and planning to monitoring and optimization. It explores how explainable AI (XAI) frameworks enhance transparency by making algorithmic decisions auditable, traceable, and intelligible to both technical and non-technical stakeholders. Case studies from the renewable energy, smart city, and green building sectors illustrate how AI tools can optimize resource usage, reduce embodied and operational carbon footprints, and enable dynamic carbon accounting models. Further, the paper analyzes regulatory landscapes and ESG mandates shaping the use of AI in sustainability contexts. It highlights global frameworks such as the EU AI Act, ISO sustainability standards, and national decarbonization strategies, detailing how compliance-driven design principles can align AI deployment with climate and governance goals. The importance of stakeholder-inclusive development, bias mitigation, and lifecycle impact analysis is emphasized as part of a broader responsible innovation agenda. Ultimately, the study proposes a multi-criteria framework for evaluating AI systems in sustainable projects, advocating for an ethically grounded and performance-driven approach that reinforces trust, accountability, and measurable environmental impact.