Navigating the AI-Sustainability Paradox

Sammanfattning

As artificial intelligence (AI) becomes embedded in business operations, it presents both promise and risks for environmental sustainability. While AI has potential to support sustainability efforts through enhanced efficiency, data optimisation, and resource management, its deployment entails substantial environmental costs, including high energy consumption, electronic waste, and reliance on critical minerals. This paradoxical tension between AI's environmental benefits and harms raises key questions for sustainability management. Despite growing academic attention, empirical insights into how organisations perceive and address this tension remain limited. This qualitative study addresses that gap by exploring how sustainability managers perceive AI's dual environmental role, and how organisations evaluate and measure it. Drawing on semi-structured interviews with sustainability managers across sectors in Sweden the study finds that while some organisations view AI's multifaceted environmental impact as critical, others do not consider it a pressing issue. Organisations do not yet measure or evaluate AI's net impact, though few are beginning to assess aspects of it. Results suggest AI's environmental implications are still emerging on the corporate agenda, requiring more deliberate integration into sustainability strategies. This research advances theory on the paradoxical tension of AI's multifaceted environmental impacts and organisational sustainability strategies, and offers practical recommendations emphasising greater awareness, enhanced evaluation practices, and collaboration both across departments and between organisations.