DSM Directive, Article 4 (text-and-data-mining exception and rights reservation)
Directive (EU) 2019/790, Article 4
In force since , under challenge in court.
A copyright and text and data mining (TDM) rule binding public and private bodies.
- Criminal exposure
- No
- Private right of action
- No
- Obligation class
- TDM
- Audit expectation
- none
- Court action
A suit challenging this law is pending. No court has blocked it. Premium data
What it requires
- Before relying on the Article 4 exception, check whether the rightholder has expressly reserved the use in an appropriate manner (Art. 4(3)); for content made publicly available online, recital 18 treats machine-readable means, including metadata and the terms and conditions of a website or service, as the appropriate way to reserve it.
- Skip or license content whose rightholder has expressly reserved the use in an appropriate manner, unless the mining involves no protected reproduction or extraction (recital 9) or another exception covers it, such as Article 3 for research organisations and cultural heritage institutions or Article 5(1) of Directive 2001/29/EC for temporary acts of reproduction.
- Mine only content you access lawfully in the first place
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
Enforcement body
Member States' courts, under the national law that transposes Article 8(2) of Directive 2001/29/EC, which lets a rightholder whose interests an infringing activity affects bring an action for damages or apply for an injunction.
Settledness
A reference for a preliminary ruling lodged on by the Budapest Környéki Törvényszék (Case C-250/25, Like Company v Google Ireland Limited) asks in its second and third questions whether training an LLM-based chatbot is an act of reproduction and, if so, whether that reproduction of lawfully accessible works falls within the exception in Article 4 of Directive (EU) 2019/790. Two further references are pending. In Case C-806/26 (Manchester United Fanatics Klub Magyarország v Google Ireland Limited), lodged on by the Pécsi Törvényszék (Pécs High Court, Hungary), the sixth question asks whether, under Article 4(3) of Directive (EU) 2019/790, a press publisher that does not expressly and appropriately assert its right of exclusion is considered to have authorised the use of its publication for the functioning of an LLM-based chatbot. In Case C-829/26 (Mobilguru Kereskedelmi és Szolgáltató Korlátolt Felelősségű Társaság v Google Ireland Limited), lodged on by the Fővárosi Törvényszék (Hungary), the referring court asks how the scope of the right of prohibition in Article 4(3) is to be interpreted, including whether press publishers can exclude the use of their works entirely or only beyond the fair dealing in Article 5(1) of Directive 2001/29/EC.
- As of
- Under challenge
- Yes
- Open questions
- Which forms of reservation (a robots.txt directive, a metadata field, a protocol signal) satisfy "in an appropriate manner, such as machine-readable means" in Article 4(3) when the Directive names no format?
- Does a reservation stated only in the natural-language terms and conditions of a website count as machine-readable means under Article 4(3), given that recital 18 lists terms and conditions among them?
- Does a reservation made after a work was reproduced remove the exception for the copies already made and retained under Article 4(2)?
- Can a website operator that is not the rightholder of the works it hosts make the reservation that Article 4(3) assigns to "their rightholders"?
- Does training a generative AI model on a work fall within the definition of "text and data mining" in Article 2(2), an automated analytical technique aimed at generating information such as patterns, trends and correlations?
- Is a work that a crawler reached by circumventing an effective technological measure under Article 6(1) of Directive 2001/29/EC "lawfully accessible" under Article 4(1)?
- Can a press publisher's reservation under Article 4(3) of Directive (EU) 2019/790 exclude the use of its press publication for text and data mining entirely, or only use that goes beyond the fair dealing in Article 5(1) of Directive 2001/29/EC to which Article 15(1) refers?
What it makes you log
Logging duty
Article 4 requires Member States to provide a text and data mining exception for lawfully accessible works, which applies only where the rightholder has not reserved the use. It requires no log or record to be kept, and its allowance to retain copies for as long as necessary concerns the mined copies, not a record.
- Kind
- None
- As of
Also on the record
EEA status
- Status
- Pending
- Joint committee decision number
- 333/2023
- Source link
- https://eur-lex.europa.eu/eli/dec/2024/1465/oj/eng
- Decision date
What this law does
Article 4 requires Member States to provide an exception or limitation for reproductions and extractions of lawfully accessible works for the purposes of text and data mining, and paragraph 2 permits retaining those copies for as long as the mining requires.
Unlike the Article 3 exception, which is confined to research organisations and cultural heritage institutions acting for scientific research, Article 4 is limited by neither purpose nor actor, and recital 18 records that mining techniques are widely used by private and public entities alike, including for the development of new applications and technologies.
Paragraph 3 makes the exception conditional on the rightholder not having expressly reserved the use in an appropriate manner, which the Article illustrates as machine-readable means for content made publicly available online, so a reservation removes the exception rather than merely signalling a preference. Article 7(1) renders contractual override unenforceable for Articles 3, 5 and 6 but does not list Article 4.
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 .