Key points are not available for this paper at this time.
The widespread use of AI technologies impacts individuals, organisations, and societies in ways that warrant more comprehensive investigation. Inequalities in benefiting from AI advancements may lead to an ‘AI divide’—a gap between those who can generate successful and sustainable outcomes with AI and those who cannot. Although human-AI interactions can lead to various outcomes, like increased efficiency and improved task performance, this study emphasises the importance of fostering sustainable outcomes that account for the social, ethical, and environmental impacts of AI usage. In doing so, this work demonstrates how achieving success in AI utilisation can also contribute to achieving the United Nations’ Sustainable Development Goals (SDGs). To better understand the disparities that divide individuals, organisations, and societies that can generate successful and sustainable outcomes with AI from those who cannot, we analyse and explain how several social, technical, and sociotechnical inequalities limit sustainable outcomes of human-AI interaction by integrating insights from a systematic literature review on AI inequalities with existing research on the digital divide and sociotechnical system theory. This approach conceptualises the AI divide as a multi-level and circular construct, linking performance-related outcomes with sustainable development goals to advance individual, economic, environmental, and societal well-being through human-centered AI.
Hammerschmidt et al. (Fri,) studied this question.