PolandUstawa o prawie autorskim i prawach pokrewnych, Text-and-Data-Mining Exceptions
Ustawa z dnia 4 lutego 1994 r. o prawie autorskim i prawach pokrewnych (tekst jednolity Dz.U. 2025 poz. 24) art. 26² i art. 26³, dodane ustawa z dnia 26 lipca 2024 r. (Dz.U. 2024 poz. 1254)
In force since .
A copyright and text and data mining (TDM) rule binding public and private bodies.
- Obligation class
- TDM
- Audit expectation
- none
What it requires
- Honor a rightsholder's express opt-out from the general text-and-data-mining exception; for a work made available so that anyone can access it at a time and place of their choosing, that opt-out is valid only when expressed in a machine-readable format together with metadata.
- Rely on the unconditional research text-and-data-mining exception, which no opt-out can defeat, only where you are a cultural-heritage institution or a covered higher-education or research body acting without a direct or indirect financial-gain purpose, and only while you keep the mined reproductions under access controls limited to authorized persons.
These lines are LexLint's reading of the law's text.
In the frameworks
- NIST AI RMF
- MAP 4.1Approaches for mapping AI technology and legal risks of its components – including the use of third-party data or software – are in place, followed, and documented, as are risks of infringement of a third party’s intellectual property or other rights.
- NIST AI 600-1
- GAI-RISK-10Intellectual Property
- GAI-RISK-12Value Chain and Component Integration
- MIT mitigations Draft
- FINOS AIGF
- AIR-PREV-007Legal and Contractual Frameworks for AI Systems
- ETSI EN 304 223
- 5.2.4-2.1Minimum record for training data: its source, such as a page URL, and when it was collected.
Our reading of where this law's duties sit in each framework. It never says that running a control meets the law. How we read a framework against the law
Identifiers refer to ETSI EN 304 223 V2.1.1 (2025-12), © ETSI 2025, all rights reserved. ETSI owns the standard; no ETSI text (titles or sentences) is reproduced here. The standard is free to download from ETSI.
Who enforces it
Settledness
- As of
- Open questions
- What counts as a reservation 'w formacie przeznaczonym do odczytu maszynowego' under Article 26³ paragraph 2 for a work made publicly available online, and does a robots.txt directive or a note in the terms of use meet it?
- Does the training of a generative model on lawfully disseminated works fall within 'eksploracji tekstów i danych' under Article 26³ paragraph 1?
What it makes you log
Logging duty
Articles 26² and 26³ allow text and data mining copies, subject to the rightholder's reservation, and limit how long copies may be stored and who may reach them. They require no log or record to be kept.
- Kind
- None
- As of
What this law does
Article 26², added by the Act of transposing the EU Digital Single Market Copyright Directive, lets a cultural-heritage institution or a body covered by Article 7(1)(1), (2), or (4) to (8) of the Law on Higher Education and Science reproduce a work to text-and-data-mine it for scientific research, without a rightsholder opt-out, so long as the activity is not carried out for direct or indirect financial gain; the reproductions may be retained only under access controls limited to authorized persons, and the rightsholder may apply only the measures necessary to keep the network or database where they are stored secure.
Article 26³ lets anyone reproduce a disseminated work for text-and-data-mining, unless the rightsholder has reserved that use. A reservation must be express, and for a work made publicly available so that anyone can access it at a place and time of their choosing, it must be in a machine-readable format together with metadata. Reproductions made under Article 26³ may be retained only as long as necessary for the mining purpose.
When LexLint raises it
When your app profile says your app crawls the web or trains models.
Back to the example · Lint your app
Page updated .