Year: 2024
Type: dissertation
Abstract: Artificial intelligence (AI) investments by organisations across sectors have skyrocketed in recent years, driven by their potential to enhance business performance and marketing effectiveness through improved decision quality in task augmentation and automation within competitive environments.However, the black-box nature of many modern AI systems, powered by machine learning (ML) algorithms, raises significant ethical and societal challenges, making responsible AI a critical and widely discussed topic.Its scope includes ethical and responsible approaches to designing, deploying, and implementing AI systems that uphold individual rights and promote social good.In response to the growing discourse, this thesis, comprising three empirical essays, adopts a pragmatic philosophical approach to examine the multifaceted implications of AI adoption across organisational, managerial, and individual levels using a multi-method research design.Essay 1, based on a practice-based view (PBV), develops and validates a conceptual model informed by corporate social responsibility (CSR) and information technology (IT) business value to examine how responsible AI practices create business value, supported by multiple case studies for empirical analysis.Essay 2 advances theoretical understanding of managing AI's evolving frontiers (e.g., autonomy, learning, and inscrutability) through a qualitative, interpretive study.It introduces a grounded model that integrates technical and social aspects of responsible AI design and governance, connecting instrumental and humanistic outcomes to highlight the multifaceted nature of responsible AI implementation in contemporary organisations.Essay 3 focuses on the consumer side, investigating how generative AI (GenAI) impacts consumer intentions and trust in healthcare marketing.Through an experimental design, it integrates construal level theory (CLT), the accessibility-diagnosticity model (ADM), and cue utilisation theory (CUT) to reveal how goal-versus duty-oriented messaging drives the GenAI-powered healthcare service adoption across three domains: prevention, diagnosis, and treatment, offering insights for digital healthcare marketing communications.Collectively, this thesis contributes novel insights into responsible AI, advancing theoretical research while providing practical strategies for organisations to implement AI ethically and effectively at multiple levels.here, supported me through my struggles and celebrated my small victories along the way.This has been a transformative journey, and finally, I have made it to the finish line.
Author: Xiong, Mengran
Institution: University of Sheffield
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7009747292