Deepfake technology and the future of digital archives: implications for authenticity, reliability and public trust

Abstract

This study aims to examine the future of digital archives in the era of deepfake technology. The study adopted a qualitative research design, drawing on the expertise of archival practitioners from public agencies. Accordingly, a semi-structured interview was selected to examine the opportunities offered by deepfake technology to digital archives and the potential risks posed by deepfakes in undermining public trust in digital archives and historical memory. Findings revealed a complex duality in how synthetic media is perceived both as a disruptive force and as a transformative innovation capable of enhancing archival trust and authenticity. When ethically employed, deepfake technologies can facilitate constructive authenticity, restore and preserve digital archives and transform synthetic reconstruction into an extension of the archival mission. However, when misused, they risk destabilizing and betraying Duranti’s foundational triad of authenticity, reliability and accuracy. For policymakers, the findings point to a pressing need for updated legal and regulatory frameworks that address the evidentiary status of digital records in the age of synthetic media. For IT and digital preservation teams, the study highlights the urgency of integrating advanced detection and verification tools capable of identifying AI-generated or manipulated content. Educators in archival science, library and information studies and information governance must revise curricula to reflect the evolving digital landscape. Although a growing body of research has examined the risks posed by deepfake and synthetic media to information integrity, the archival and records management discipline remains underexplored in this discourse. In addition, while AI applications in libraries and archives are widely discussed, scholarship often emphasizes efficiency and access rather than resilience against manipulative content. Existing studies have largely concentrated on the implications of deepfakes for journalism, national security and public misinformation (Chesney and Citron, 2019), with comparatively little empirical attention to how such technologies affect archival principles of authenticity, reliability and integrity. This gap underscores the need to investigate the opportunities offered by deepfake technology to digital archives and the potential risks posed by generative AI in undermining public trust in digital archives and historical memory.