Identifying artificial intelligence use cases: toward a method that facilitates garbage can innovation processes

While artificial intelligence (AI) technologies offer much potential for a wide range of AI uses in organizations, identifying use cases involves making many decisions amidst influencing complications. While nascent methods for identifying such AI use cases seem promising, we know very little about their proven efficacy to actually guide decisions during the use case identification process. To investigate this efficacy, we conducted action design research at EnBW, one of Europe’s largest energy suppliers. We draw on interviews and an intervention at EnBW’s practices to develop a method, derive six design principles, and describe factors that increase our method’s efficacy. We therefore contribute a novel theoretical perspective on decision-making in contexts of organized anarchy, explaining how organizations can navigate the complexities of decisions in AI use case identification.