Identifying barriers to increased AI in the insurance industry : A case study

Sammanfattning

The current technological advances within the field of artificial intelligence (AI) have resulted in an increasing focus from corporations to find ways to incorporate AI into their businesses. Within the insurance industry, previous research has identified several potential use cases ranging from intelligent chatbots to fully automatic claims handling and fraud detection. However, there is a clear discrepancy between organizational ambition and practical execution which indicates that the process of implementing AI is considerably more difficult than anticipated. This study investigates the barriers associated with implementing AI by conducting a case study at Gjensidige Sverige, a Swedish insurance company evaluating an increased focus on AI. Based on observations of the corporation as a participant in AI-related projects for a year, as well as conducting interviews with employees within the organization, this thesis identifies several barriers to working with AI in insurance. A distinction between ”hard” and ”soft” barriers is made, where hard barriers consist of barriers that completely prevent the implementation of AI systems while soft barriers affect the created value of an implemented solution. This thesis argues that soft barriers needs more attention to mitigate the risk of unsuccessful AI projects.