Supplementary Appendix A: Diella Media Corpus (n = 50)

Supplementary Appendix A to Cavallo, V., & Lapenna, M., "Performing Algorithmic Integrity: Media Framing and Public Interpretation of Diella" (manuscript submitted to Internet Policy Review). The dataset is an inventory of the 50 online media texts analysed in the study, published between 11 September 2025 and 8 February 2026, covering Diella, the AI system designated as a virtual minister by the Albanian government. For each item it records an identifier, publication date, outlet, country or region, language, genre, title, the dominant frame emphasis assigned during coding, and the URL as accessed. Frame labels were developed inductively; the FCET model served as an organising lens rather than an exhaustive coding protocol. Because the labels record dominance rather than presence, they support description of the corpus but not frequency claims about individual frame categories. Two files are provided: a formatted document (.docx) and a machine-readable version (.csv). Interview transcripts from the companion fieldwork are not included: participants discussed institutional corruption and political distrust, and confidentiality was a condition of informed consent.