"Everything Everywhere All at Once" AI-Generated Scholarly Objects, Metadata, Rights, and Deposit Decisions

Abstract

This presentation was prepared for the 2026 Midwest Data Librarian Symposium (MDLS). It examines emerging questions for institutional repositories as generative AI becomes integrated into research workflows and AI-generated or AI-assisted materials increasingly become potential repository deposits. Topics include provenance and metadata for AI-assisted research objects; distinctions among human creators, AI tools, depositors, contributors, and source-material rights holders; copyright, licensing, and restricted source materials; documentation of human review and editing; and practical approaches to repository intake and review. The presentation uses an AI-generated podcast as an opening case and introduces a Publish, Revise, Pause, or Refer framework for evaluating AI-assisted deposits. AI use disclosure: Generative AI tools were used during development of this presentation to support research organization, synthesis of research notes, refinement of language and presentation structure, and the creation or refinement of selected visual materials. The author selected the sources, developed the argument and interpretations, reviewed and verified substantive claims and citations, revised the resulting materials, and retains responsibility for the final content. A Master Research Bibliography is included with this record as a separate PDF documenting the sources consulted in developing the presentation.