Working paper
Institutional Guidance on AI Transcription for Research Interviews: A Survey of 130 Universities in the United States, United Kingdom and Canada
Ankur Shrestha, Pepys, United States · 2026-10-31
Abstract
Researchers routinely upload interview recordings to cloud AI transcription tools, but it is unclear what their institutions tell them to do. We surveyed the public guidance of 130 institutions: 100 US R1 universities drawn at random from 187, 15 of the 24 UK Russell Group members, and all 15 Canadian U15 members. Each institution was coded twice, independently, in a language-model-assisted workflow from saved page text, and disagreements were resolved against the page text. Ninety institutions (69.2%, 95% CI 60.8–76.5) published something that touches AI tools and recordings or transcription, but only 73 (56.2%) gave transcription-specific guidance; 40 (30.8%) had none we could find. Sixty-one institutions name at least one tool as approved or prohibited. Zoom's AI features are approved at 41 institutions and Microsoft Teams at 20, while stand-alone meeting bots are barred: Read.ai at 17, Fireflies.ai at 11 and Otter.ai at 20, yet Otter.ai is also approved at 5. Only 25.6% of institutions with guidance say researchers must disclose AI transcription in the ethics protocol, and 36.7% publish no date. Inter-coder agreement was high on whether guidance exists (κ = 0.89) and moderate on stance (κ = 0.68). Institutions give researchers different answers about the same products, and much of the guidance is drawn from IT pages on meetings rather than from ethics offices.
130
institutions surveyed (US R1, UK Russell Group, Canada U15)
56%
give transcription-specific guidance
20 vs 5
institutions prohibiting vs approving Otter.ai, as of 30 September 2026
26%
of the 90 institutions with guidance say researchers must disclose AI transcription in the ethics protocol; the rest are silent or unclear, so this is a lower bound
What transcription software do universities allow?
It depends on the university. Policies vary, and the same tool can be approved at some institutions and prohibited at others. In the public guidance of 130 large research universities, collected on 30 September 2026, the tools most often approved were platforms the institution already licenses, such as Zoom's AI features (approved at 41) and Microsoft Teams (20), while stand-alone meeting bots were the tools most often barred. Otter.ai was prohibited at 20 institutions and approved at 5.
Sixty-one of the 130 institutions name at least one tool as approved or prohibited, and 40 had no guidance we could find. Counts are institutions, not mentions, and a tool can be approved at one institution and prohibited at another.
| Tool family | Institutions approving | Institutions prohibiting |
|---|---|---|
| Zoom (AI Companion and built-in transcription) | 41 | 1 |
| Microsoft Teams (including Teams Premium) | 20 | 0 |
| Microsoft 365 (Word Transcribe, Copilot, Stream) | 13 | 1 |
| Whisper-based or local tools | 12 | 1 |
| Rev | 6 | 3 |
| Otter.ai | 5 | 20 |
| Read.ai | 0 | 17 |
| Fireflies.ai | 1 | 11 |
The sample covers large research universities in the United States, United Kingdom and Canada, and pages change, so check your own ethics office or IRB before you upload research recordings; a missing rule is a question to ask, not a permission. No institution in the sample names Pepys, and the study did not test any vendor's product.
Where to find it
- SSRN preprint: View on SSRN
- DOI: 10.5281/zenodo.23846598 (Zenodo preprint)
- Dataset DOI: 10.5281/zenodo.23846481 (CC BY 4.0)
Putting it into practice
- Workflow for researchers: research interview transcription
- Method guide: qualitative research transcription
- Group recordings: how to transcribe a focus group
- Thesis work: transcribing an interview for a dissertation
- Wider context: transcription in academic research
- Plain-language summary: what universities actually allow
- Before you upload: an eight-question checklist
Cite
@misc{shrestha2026institutional,
author = {Shrestha, Ankur},
title = {Institutional Guidance on AI Transcription for Research Interviews: A Survey of 130 Universities in the United States, United Kingdom and Canada},
year = {2026},
howpublished = {SSRN preprint, abstract\_id 7685681},
doi = {10.5281/zenodo.23846598},
url = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7685681}
}RIS
TY - RPRT AU - Shrestha, Ankur TI - Institutional Guidance on AI Transcription for Research Interviews: A Survey of 130 Universities in the United States, United Kingdom and Canada PY - 2026 DA - 2026/10/31 PB - SSRN ET - 1.0 UR - https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7685681 DO - 10.5281/zenodo.23846598 N1 - Dataset: https://doi.org/10.5281/zenodo.23846481 KW - AI transcription KW - research ethics KW - institutional review boards KW - data governance KW - qualitative research KW - document analysis AB - Researchers routinely upload interview recordings to cloud AI transcription tools, but it is unclear what their institutions tell them to do. We surveyed the public guidance of 130 institutions: 100 US R1 universities drawn at random from 187, 15 of the 24 UK Russell Group members, and all 15 Canadian U15 members. Each institution was coded twice, independently, in a language-model-assisted workflow from saved page text, and disagreements were resolved against the page text. Ninety institutions (69.2%, 95% CI 60.8–76.5) published something that touches AI tools and recordings or transcription, but only 73 (56.2%) gave transcription-specific guidance; 40 (30.8%) had none we could find. Sixty-one institutions name at least one tool as approved or prohibited. Zoom's AI features are approved at 41 institutions and Microsoft Teams at 20, while stand-alone meeting bots are barred: Read.ai at 17, Fireflies.ai at 11 and Otter.ai at 20, yet Otter.ai is also approved at 5. Only 25.6% of institutions with guidance say researchers must disclose AI transcription in the ethics protocol, and 36.7% publish no date. Inter-coder agreement was high on whether guidance exists (κ = 0.89) and moderate on stance (κ = 0.68). Institutions give researchers different answers about the same products, and much of the guidance is drawn from IT pages on meetings rather than from ethics offices. ER -
Version history
- v1.0 (2026-10-31): First posted.