Short answer: You can share qualitative interview data, but rarely as an open download. Plan it before the first interview. Get consent for reuse as well as for participation, pseudonymize as you transcribe, test whether a transcript could still single someone out, and deposit it under controlled access. If the data cannot be shared safely, share the interview guide, codebook and metadata instead.
Funders and journals increasingly expect qualitative data sharing, but interview transcripts are among the hardest research data to release safely. Each one is a long, detailed account of one person's life. 2026 has also changed the European rules. The European Data Protection Board (EDPB) released draft guidelines on scientific research in April and draft guidelines on anonymisation in July. This guide explains what those documents mean in practice for PhD students, research labs and research firms planning to archive or share interview data, and gives you a decision matrix to use in your data management plan.
\n\nThis article is methodological guidance, not legal advice. Confirm decisions with your data protection officer, ethics committee and target repository.
\n\nTwo EDPB documents now shape how European researchers should think about qualitative data sharing. Both were still drafts as of October 12, 2026.
\n\nOn April 16, 2026, the EDPB announced draft guidelines on processing personal data for scientific research. They list six indicative factors for what counts as scientific research, including a systematic approach, ethical standards, and verifiability and transparency. They also state that further processing for research is presumed compatible with the original purpose. They accept \"broad consent\" where research purposes cannot be fully specified at collection, provided safeguards compensate, and they also allow \"dynamic consent\", where participants agree to each project as its purpose becomes clear. The public consultation ran until June 25, 2026. We could not find a final version as of October 12, 2026.
\n\nThe EDPB adopted draft Guidelines 02/2026 on Anonymisation on July 7, 2026, for public consultation. According to a Sidley Austin summary, the consultation closes on October 30, 2026, and the draft is intended to replace the 2014 Article 29 Working Party opinion on anonymisation techniques. The draft says anonymity must be assessed from the perspective of each party that will receive the data, using \"means reasonably likely to be used\". It proposes three tests:
\nThe draft notes that re-identification is generally more likely to succeed against record-level data with high dimensionality and high resolution. That is our reading of why interview transcripts are hard to anonymize. A single transcript can contain a job title, a town, a family situation and a rare event, and together they may point to one person even after the names are gone.
\n\nNo. Replacing names with codes while you keep a key file is pseudonymization, and pseudonymized data remains personal data under the GDPR. Article 89 GDPR names pseudonymization as one of the safeguards for research processing. It is a safeguard, not an exit from the regulation. A transcript leaves the GDPR's scope only if it passes an anonymity assessment from the perspective of the people who will receive it.
\nMany qualitative studies cannot pass that test without stripping out the context that made the data worth collecting. That is the core tension in qualitative data sharing, and it is why the usual answer is controlled access rather than open release.
\n\nThe UK Data Service warns that anonymisation should not be a single step left to the end of a project. It recommends combining three elements: transparency with participants, technical protection such as pseudonymization, and governance through access controls. One element does not replace another. The workflow below puts that advice in order.
\nThis matrix is an original Qualitati planning tool. Use it in your data management plan to match each data type to a realistic sharing route. It is a starting point for a discussion with your repository and data protection officer, not a compliance ruling.
\n| Data and context | Typical re-identification risk | Suggested route | What to share at minimum |
|---|---|---|---|
| Short, low-detail transcripts on a non-sensitive topic, consent for reuse obtained | Low to moderate | Safeguarded access (registered users, end-user licence) | Pseudonymized transcripts, codebook, interview guide |
| In-depth interviews with rich life-history detail | High | Controlled access with an application and approval | Generalized transcripts, data dictionary, metadata |
| Small or easily recognizable population (one organization, rare condition, elite interviewees) | Very high | Controlled access or secure environment, often embargoed | Metadata, interview guide, codebook, aggregate theme tables |
| Raw audio or video recordings | Very high (voice and face are identifiers) | Share only with explicit consent, under controlled access | Transcripts instead of recordings |
| No consent for reuse, or sensitive data from vulnerable participants | Not shareable | Do not deposit the data | Metadata, interview guide, codebook, analysis audit trail |
Who this is for: doctoral researchers writing a data management plan, principal investigators on EU-funded or UK-funded projects, and research firms whose clients ask about data retention and reuse.
\nWhen not to share: skip data deposit when consent did not cover reuse, when participants are few enough to recognize one another, or when anonymization would remove the meaning the data was collected for. Commercial studies under client confidentiality usually fall into this group too. In those cases, share the research materials and the analysis audit trail instead.
\n\nQualitati is a European qualitative research platform for universities and research firms. It runs AI-moderated interviews, focus groups and conversational surveys, and it codes the results with ThemeLens and the QDA Workspace. Several of its features are relevant to qualitative data sharing:
\nQualitati does not decide whether your data is anonymous or whether your consent covers reuse. Those remain the researcher's judgment, made with the ethics committee and the data protection officer.
\n\nYes, if you have a lawful basis, participants were informed about reuse, and you apply safeguards such as pseudonymization and controlled access. Transcripts that pass an anonymity assessment fall outside the GDPR, but many detailed qualitative transcripts cannot pass one.
\nYes. While anyone can link the codes back to people, for example through a key file, the data remains personal data. Pseudonymization reduces risk and is listed as a research safeguard in Article 89 GDPR.
\nThe July 2026 draft Guidelines 02/2026 propose no record isolation, no linkage and no inference, assessed from the perspective of each party that will receive the data, using means reasonably likely to be used.
\nUsually not. Seek advice from your data protection officer and ethics committee. The safer option is to share metadata, the interview guide and the codebook rather than the data itself.
\nBroad consent covers future research whose exact purposes are unknown at collection. The EDPB's 2026 draft research guidelines accept it when ethical standards and additional safeguards compensate for the less specific purpose.
\n\nQualitative data sharing works when it is designed in from the start: granular consent, pseudonymization during transcription, an honest anonymity test against the EDPB's three criteria, and controlled access by default. If the data cannot be shared safely, sharing the materials still makes your study more transparent. Start free with 30 monthly credits to run your next interview study on an EU-based platform, or view transparent pricing. For broader planning, see how to write a research plan.
\n\nSources: EDPB press release, April 16, 2026; EDPB Guidelines 02/2026 on Anonymisation (draft); Article 89 GDPR; UK Data Service; CESSDA Data Management Expert Guide.
\n\nLast updated: October 12, 2026. This article is general methodological guidance, not legal advice. A human editor should review jurisdiction-specific claims before reuse.
\n","related":[{"slug":"how-to-write-research-readout-2026","title":"How to Write a Research Readout Stakeholders Act On","description":"A 2026 guide to research readouts: the evidence chain from finding to decision ask, a readiness checklist, format-by-decision matrix, and where AI helps.","category":"Research Operations","date":"2026-09-20","readTime":"10 min read","author":"Qualitati Research Team"},{"slug":"ux-research-capacity-gap-2026","title":"The 2026 UX Research Capacity Gap","description":"Research demand hit 66% in 2026 while support stayed flat. A triage matrix for routing requests between AI, self-serve studies, and researcher-led work.","category":"Research Operations","date":"2026-09-06","readTime":"9 min read","author":"Qualitati Research Team"},{"slug":"screening-fraudulent-participants-ai-interviews-2026","title":"Fraud Detection in AI-Moderated Interviews (2026)","description":"Can AI-moderated interviews detect fraudulent participants? The tools, quality signals and consent rules that apply in 2026, plus a three-layer fraud screen.","category":"Research Operations","date":"2026-09-05","readTime":"8 min read","author":"Qualitati Research Team"}]}