Diversity Governance: Gestaltung eines Diversitätssensiblen Ansatzes zur Automatisierten Identifikation und Reduktion Impliziter Diskriminierung in der Unternehmenskommunikation

Diversity programs aim not only to unlock economic potential, but also to support companies’ development toward a contemporary and ethical corporate culture. Despite companies’ positive intentions to implement diversity programs, recent data indicate slow progress in this area. Our study presents a novel and innovative approach to promoting diversity in business that aims to reduce implicit discrimination within corporate communications. Similar to corporate governance, the diversity governance we propose seeks to shape the actions of employees and managers in ways that empower them to communicate in accordance with diversity principles. In this study, we present results related to the design and technical implementation of diversity governance software designed to automatically analyze corporate communication data, for example in the form of texts and images, for implicit discrimination. An automated assessment of corporate communications through artificial intelligence offers advantages in terms of speed and consistency, unaffected by human limitations such as fatigue in the case of extensive analysis input. Beyond assessment, the software focuses on enhancing the analyzed text and images by providing AI-based recommendations, such as alternative phrases. To gather feedback on the design and technical implementation of our software, we conducted twelve interviews with experts in the field of diversity. Finally, we used our findings to derive practical recommendations for the successful implementation of diversity governance.