Artificial Intelligence in Organizations: Managing the Lifecycle of Conversational Agents

Abstract

Given the rapid advances of information technology, led by increasing dissemination and possibilities, many areas of science, society and the economy are being impacted upon. In this context, artificial intelligence (AI) has developed into a disruptive technology. Driven by the large availability of data, increased computing power, and algorithmic and technological advances, AI is a decisive factor in the digital transformation of organizations. As a result, companies are spendingsignificant resources on AI-based systems and related applications, e.g., to improve customer engagement and remain competitive.

Conversational Agents (CAs), such as chatbots, are an emerging natural language-based application of AI, which provide organizations with an increasingly intelligent, social, learning, integrative, and cost-effective platform to support and automate the delivery of services and information. CAs offer new forms of scalability and availability (24/7), speed, and personalization. However, despite their possibilities, organizations are struggling to realize the full potential of CAs in real-world environments. AI-based CAs pose new challenges due to their unique characteristics. They are not sufficiently represented in current research, which leads to many open questions and opportunities regarding the management of CAs’ lifecycle, especially in regard to different activities in the introduction, design, implementation, operation, and improvement phases.

From a broader, organization-wide perspective, the management of AI is an essential task inrealizing the new value propositions. However, knowledge on how to structure AI transformations and effectively manage specific applications is limited, leading to a high failure rate of AI initiatives. There is a need for research to support organizations in systematically developing AI (implementation) competencies in order to master all the complexity of this new technology. By switching viewpoints between organization-wide and system-specific perspectives, this dissertation contributes to the emerging field of (1) AI in organizations in general by providing guidance for navigating, managing, and (re)evaluating AI transformations, and (2) CAs in particular, by providing various forms of validated design knowledge for their lifecycle management. Together, these themes facilitate implications for a broader understanding of AI management in organizational settings.