AI-Driven Brand Communication: Power, Trust and Identity in Digital Marketing Ecosystems

Abstract

Artificial intelligence (AI) is reshaping digital brand communication by redistributing communicative power among brands, platforms, algorithms and consumers. This review critically examines how AI-mediated tools influence power, trust and identity within digital marketing ecosystems. It focuses on applications such as chatbots, recommender systems, generative content tools, sentiment analysis and algorithmic personalisation, and considers their implications for consumer-brand relationships. A structured narrative synthesis approach was used to review multidisciplinary literature and selected case-based evidence published mainly between 2018 and 2025, while retaining earlier foundational studies relevant to branding, trust and digital communication. Thirteen case studies were thematically organised into four domains: AI-powered trust formation, brand identity evolution, algorithmic power in platform-mediated branding, and AI-driven personalisation and engagement. The synthesis indicates that AI can support timely, scalable and personalised communication, improve service responsiveness and strengthen some aspects of brand relevance and consumer engagement. However, these benefits are accompanied by concerns about data privacy, algorithmic opacity, uneven service quality, loss of human agency and the possible fragmentation of brand identity. The review also shows that trust in AI-mediated brand communication depends not only on system performance but also on transparency, perceived fairness, ethical governance and the credibility of the parent brand or platform. Overall, AI should be understood as a socio-technical participant in brand ecosystems rather than as a neutral communication tool. The paper contributes by integrating power, trust and identity into a conceptual account of AI-driven brand communication and by identifying areas for future empirical research on responsible and coherent AI use in digital marketing.