When a data set was assembled by an intern: what to do first
Secure the raw files, the assembly scripts and the intern's status confirmation before anything else moves. Ownership of an intern-assembled data set follows the internship agreement first and the surrounding facts second: who gave the instructions, whose equipment was used, and whether the intern acted on behalf of the organisation. A dated record of these facts is the first document any dispute will ask for.
The legal basis
European Union law sets a common framework for treating an organised collection of data as a database under Directive 96/9/EC. Within that framework, two layers of protection can apply at once: copyright in the selection and arrangement of the contents, and a separate right that rewards the investment made to obtain, verify or present them. Ownership of each layer typically follows whoever actually performed the qualifying acts, subject to national implementing rules and to any contract that assigns the result elsewhere.
An internship agreement that is silent on this point does not resolve the question by itself. It shifts the enquiry to facts: instructions received, resources supplied, and whether the intern's contribution was directed and paid for by the organisation.
What changes in practice
Without a written assignment, the organisation cannot assume it holds either layer of protection. The gap matters most where the tooling used to build the set sits on a third-party platform: exit terms in that contract decide whether the underlying files can even be extracted once the relationship ends. A migration is blocked by an exit clause technology walks through that separate problem.
A second practical effect: any licence covering third-party data folded into the set carries its own term and its own restrictions on redistribution. Those terms run independently of the intern's status and do not pause while ownership is being sorted out.
What to do first
- Freeze the file set. The person who supervised the intern copies the raw data, the assembly scripts and any version history to a location the organisation controls, before access is revoked.
- Pull the internship agreement. HR or the signing manager checks whether it contains an assignment clause covering work product, and whether that clause names data specifically.
- Identify third-party inputs. Whoever built the set lists every external source and licence used, with the terms attached, not summarised.
- Log the instructions given. The supervisor records, in writing, what the intern was asked to do and with what resources — this becomes the evidence of direction if the agreement is silent.
- Route the file to legal review. Internal counsel or an external adviser assesses whether the current record supports the organisation's claim before the data set is used commercially.
Who decides, and on what evidence
Ownership disputes of this kind are resolved first inside the organisation, by whoever holds signing authority over IP matters, on the strength of the internship agreement and the supervision record. If the intern or a third party contests the outcome, the matter moves to the competent civil court of the applicable jurisdiction. The court examines the same three items: the agreement's wording, the instructions actually given, and who performed the acts of collection, verification and arrangement.
| Element | Why it matters | Who typically holds it |
|---|---|---|
| Raw source files and licences | Shows where each record originated and under what terms | Whoever negotiated or received the licence |
| Assembly scripts and logs | Proves the acts of extraction, cleaning and arrangement | The person who ran them — often the intern |
| Internship agreement | States, or fails to state, who owns the work product | Signed by both intern and organisation |
| Supervision record | Shows the work was directed and resourced by the organisation | Kept by the supervising manager |
Common forks
If the internship agreement assigns work product to the organisation in clear terms, the analysis stops there: the data set belongs to the organisation, subject to any third-party licences embedded in it.
If the agreement is silent and the intern worked largely on personal equipment with no supervision record, the organisation's claim is weaker and needs to rest on other evidence — payment, direction, or the intern's own admissions.
If part of the set consists of data held under licence from a third party rather than collected directly, the position resembles a separate question addressed in What to do first in this situation explained: control and ownership are not the same thing, and the licence terms decide what can actually be done with the data.
What this page does not cover
This page does not value the data set, does not draft the assignment clause, and does not resolve a dispute already in litigation. It also does not confirm chain of title for a specific jurisdiction — that requires a file built on the actual records held.
Where ownership needs to be established as a documented, provable fact for a counterparty or a court in Denmark, the relevant groundwork is described in Asset in Denmark in practice.
Where the dispute instead concerns a party who owes rather than owns, comparable groundwork appears in Debtor in Denmark and the evidence.
Related situations
A neighbouring provenance problem — where the data itself carries no record of where it came from — is covered in When the training corpus has no provenance in practice.
Sources and verification
Checked and current as of 30 May 2025. The legal basis referenced above is Directive 96/9/EC on the legal protection of databases. Where a data set also embeds original software components, Directive 2009/24/EC on the legal protection of computer programs may apply in parallel to those components.
FAQ
What is the first step when a data set was assembled by an intern?
Freeze and copy the raw files, scripts and version history to organisation-controlled storage before access to the intern's accounts is revoked.
Which deadline is already running?
The internship term and any notice period set out in the agreement run regardless of the ownership question. Where third-party data was folded into the set, the licence's own term runs in parallel and does not pause for an internal review.
Who decides, and on what evidence?
Internally, whoever holds signing authority over IP matters decides first, on the agreement and the supervision record. If contested, the competent civil court decides on the same evidence, plus the acts of collection actually performed.
A first response on a matter like this arrives within minutes during business hours; a price for the review follows within one business day.