Institutional Guidance on AI Transcription and Recorded Research Interviews: 130-University Dataset

Coded dataset behind the working paper "Institutional Guidance on AI Transcription for Research Interviews: A Survey of 130 Universities in the United States, United Kingdom and Canada" (Shrestha, 2026). It records what the public web pages of 130 research universities (100 US R1 universities drawn at random from 187, 15 of the 24 UK Russell Group members, and all 15 Canadian U15 members) say about using AI tools to transcribe recorded research interviews, as saved on 30 September 2026. Start with release/: DATASET.csv is the one-row-per-institution table, DATASET_TOOLS.csv the tool-level table, with codebook, methodology, tables and a recalculation of every headline figure (python3 release/test_headlines.py). The rest of the package is the audit trail: both coders' codings, adjudication notes, page URLs, capture times, SHA-256 hashes, short verbatim quotes, the study plan with its hash log, and the analysis scripts. Raw page text is third-party material and is not redistributed; hashes and URLs allow verification. Competing interest: the author is affiliated with Pepys, which sells an AI transcription service. Pepys is not among the tools coded, and the dataset does not compare, rank or recommend products. Coding was language-model-assisted under a written codebook; see the paper for limitations. The study was not registered with a public registry.