Artificial Intelligence (AI) & Decision-making: Will AI trustworthiness facilitate progress towards sustainability?

Sebian, Ben; Ottogalli, Marta

Abstract

Decision-makers, from governments to the private sector, are pressured to tackle global sustainability challenges. Applications of artificial intelligence (AI) aim to help decision-makers by simplifying access to a large amount of information, addressing data gaps, improving decisions’ efficiency, automating burdensome processes and more. Humans have been reluctant to adopt AI to take decisions, even when, in some cases, it has proven to outperform humans’ judgement. Hence, the research aims to investigate the influencing factors that generate algorithm aversion among decision-makers during decision-making processes aimed at addressing sustainability challenges. Through a series of mixed research methods, participants including governments, the private sector, and intergovernmental organisations (IGOs) took part in the preliminary research, by completing a scenario-based survey and semi-structured interviews to further investigate algorithm aversion during sustainability decision- making. It must be noted that a multi-stakeholder approach is necessary to coordinate groups of stakeholders that are required to drive sustainable progress. This preliminary study showed that there is algorithm aversion and as a result: i) Decision-makers are not willing to use AI cost-benefit analysis models that they do not understand to evaluate decisions; ii) Decision-makers are more interested in using AI technologies that are based on factors they have helped co-design; iii) Decision-makers are mostly interested in automating secondary tasks that can help identify the correct data and information to take a decision effectively; iv) Decision-makers are interested in using AI that helps them identify knowledge that can directly support their decisions.

Citation

Sebian, B., & Ottogalli, M. (in press). Artificial Intelligence (AI) & Decision-making: Will AI trustworthiness facilitate progress towards sustainability?. . Springer

• Deposit Date: Jul 16, 2024
• Publisher: Springer
• Peer Reviewed: Peer Reviewed
• Keywords: Artificial intelligence, Sustainability, Decision-making, Risk, Trust, Algorithm aversion
• Contract Date: Jun 6, 2024