AI-driven corporate reputation measurement in digital ecosystems: A systematic literature review

Traditional corporate reputation measurement relies on surveys, interviews, and expert evaluations. These methods are limited by subjectivity, labour-intensive processes, and inability to capture rapidly evolving stakeholder perceptions. Biases such as recall and framing further weaken reliability, underscoring the need for data-driven alternatives. AI offers a transformative approach, enabling automated, real-time analysis of large-scale, unstructured data. However, comprehensive reviews of AI applications in this field are lacking. This study systematically reviews 104 studies (2000-2024) using PRISMA guidelines to examine how AI and machine learning enhance corporate reputation measurement. It explores key AI methods, theoretical foundations, global collaborations, and emerging challenges. The review identifies critical theoretical frameworks, including Consumer Information Processing Theory, Cognitive Theory, and Social Identity Theory, which influence algorithmic design in reputation management. Data sources range from simulation data and surveys to social media platforms, though reliance on simulation data reveals practical application gaps. Four challenges dominate: narrow AI scope, scalability issues, data complexities, and ethical concerns. The research reveals major gaps in visual data use, with only 2 % of studies employing image analysis. AI mechanisms often assess product-level reputation rather than organisational reputation. As the first systematic review of AI-driven corporate reputation measurement, it emphasizes the need for standardized terminology, integration of trust-reputation dynamics, and inclusion of emerging domains like social commerce and dynamic reputation metrics. It identifies issues such as narrow product-level focus, subjective feedback, and scalability constraints. It proposes an integrated roadmap combining traditional methods with AI tools to advance next-generation reputation measurement.