Policymakers generate digital records on a daily basis, a portion of which is preserved in archival repositories. However, access to these archival resources remains severely restricted for a multitude of reasons, including data protection, information sensitivity, national security concerns, and intellectual property rights. Artificial Intelligence (AI) holds significant potential for enhancing access to archives, yet its application in this domain remains largely experimental. Beyond the skill gaps that contribute to the continued obscurity of archives, it is imperative to examine issues of distrust and misunderstanding. This paper argues that although public officials, archivists, and academics are guided by similar professional principles—often codified in their respective ethical codes—these principles are seldom shared or transparently communicated across these professional boundaries. This lack of dialogue fosters a climate of mutual distrust among stakeholders. Furthermore, a distinct distrust in the technology itself presents a significant barrier to the effective implementation of AI tools. Consequently, the authors propose that the explicit articulation and discovery of a shared professional ethic among stakeholders can serve as a foundation for deeper human collaboration. Such collaboration, in turn, is posited to catalyze the development of trust in AI systems and tools. This research is grounded in semi-structured interviews with thirty experts, including public officials, archivists, historians, digital humanities scholars, and computer scientists. Previous research has predominantly focused on the preservation of digital records rather than on access to them and has centered more on the role of archivists than on that of records producers, such as public officials. This article represents a pioneering effort to examine the application of AI in digital archives through the critical lenses of trust and collaboration across the entire records continuum.