Interview notes should enter an AI-assisted workflow only after you have resolved the relevant permission and sensitivity questions. Consent to an interview does not, by itself, answer whether an external tool may process the notes or whether particular statements may be published.
Keep three permission questions separate
As editorial review questions, distinguish:
- Collection: What did the participant understand about recording or note-taking?
- Tool use: Was the proposed AI-assisted processing explained, including the material involved?
- Publication: Which statements may appear, with what attribution and review arrangements?
These are separate worksheet entries, not interchangeable tick boxes. An interviewee may agree to background discussion without agreeing to quotation. This article offers review questions, not legal conclusions about a particular interview.
Do not use customer or private interview material in the exercise below. Synthetic notes are sufficient to design the workflow.
Understand the limited role of data controls
OpenAI says sharing for model improvement is enabled by default on personal Free, Plus and Pro workspaces. Turning off “Improve the model for everyone” means new conversations are not used to train models and can remain in chat history.
That setting addresses the documented training use. It does not establish what an interviewee agreed to, what an employer permits or whether a quotation is suitable for publication. Resolve those questions independently rather than treating a setting as consent.
Build an interview-use worksheet
For each proposed excerpt, record:
- Material identifier and intended drafting purpose.
- Collection understanding and where it was recorded.
- Status of tool-use disclosure and permission.
- Publication status: attributed, anonymous, background-only or unresolved.
- Details unnecessary for the task.
- Required clarification and responsible reviewer.
Use “unresolved” when the record is incomplete. Removing a name is not enough for your editorial review if the remaining details still make the speaker recognisable to the intended audience.
Hypothetical drafting decision
Imagine synthetic notes from a fictional student interview in Chennai. One line describes commuting difficulties, another names a fictional lecturer and a third is marked “background only”.
The worksheet says collection was agreed, tool use was not discussed and no quotations were approved. The correct next step is not to submit the notes. Instead, practise structure with a newly written fictional paragraph about a generic commute, and separately clarify the outstanding permissions.
Evaluate a later draft by tracing every proposed quotation to its approved status. Stop publication if attribution or intended use remains ambiguous.
If researching access, check the AI tools catalog, inspect access format and save terms before buying. Marketplace account/code/token offers are not proof of official subscription equivalence.
Sources and scope
Source material checked 29 October 2026; features/policies may change. OpenAI’s data-controls documentation establishes the stated training-control effect. The permission worksheet and synthetic interview decision are editorial advice, not a legal assessment or a claim that data controls settle consent.