Abstract

The rapid adoption of AI-driven recommendation systems in New Zealand’s digital media industry has transformed how audiences can access news, while simultaneously contributing to declining public trust. This study examines how explainable artificial intelligence, supported by data ethics and governance frameworks, can be used to rebuild trust in algorithmically curated news environments. A qualitative, desk-based research design was employed, using a systematic literature review of 38 academic and policy sources selected from an initial pool of 60 articles. Thematic synthesis identified four interconnected problem areas: declining trust in news accuracy, algorithmic opacity resulting in limited user comprehension, risks to fairness and editorial independence, and unresolved privacy and accountability concerns. The findings demonstrated that trust in AI-driven media systems is conditional on transparency, fairness, user control, and responsible data governance. This study proposes an integrated framework for trustworthy AI that aligns explainability with user-centred design, editorial practices, and culturally grounded data governance, including M\={a}ori sovereignty principles. The research contributes to both theory and practice by offering context-specific recommendations for New Zealand media organisations to deliver trustworthy personalised news systems while maintaining their democratic role in society.

Keywords:
Explainable AI; News Recommender Systems; Media Trust; Algorithmic Transparency; Data Governance.

Download the article (pdf)