Copyright Foundations for Building Large Language Models in the Slovenian Legal System: Lecture by Dr. Maja Bogataj Jančič

In the introduction, Dr. Maja Bogataj Jančič highlighted the key question she would go on to address: “Is it necessary to obtain consent from authors or rightsholders in order to train devices on copyright works?”

Lawmakers around the world have approached this issue in different ways. The European legislator introduced two new exceptions for text and data mining (Articles 3 and 4 of Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market – the so-called CDSM Directive). Article 3 of the CDSM Directive establishes an exception for text and data mining for the purposes of scientific research, while Article 4 of the CDSM Directive provides a “general” exception for all other purposes, including commercial use.

These two exceptions were transposed into Slovenian law through an amendment to the Copyright and Related Rights Act (Slo. Zakon o avtorski in sorodnih pravicah – ZASP), which now contains two corresponding exceptions to exclusive rights: Article 57.b of ZASP governs the exception for text and data mining for scientific research purposes, while Article 57.a ZASP applies to all other purposes, including commercial ones.

Both Slovenian exceptions allow the free reproduction of lawfully accessed works for the purpose of text and data mining. They also permit the digitisation of analogue content and remote access to content, however only for text and data mining (TDM) purposes. In the case of the scientific research exception (Article 57b ZASP), the sharing and making available to the public of the results of TDM is also permitted. Rights holders are obliged to ensure that beneficiaries of both exceptions can effectively perform TDM. If they fail to do so, they must take action within 72 hours, or they may face sanctions.

She then explained the notion of “lawful access” under ZASP. As she noted, the concept is defined more narrowly in Slovenian law than in Recital 14 of the CDSM Directive. Under ZASP, lawful access includes access based on open licences, contracts, or other legal bases such as exceptions and limitations to copyright or special legal provisions. However, it does not explicitly refer to access to content that is freely available online. This issue was raised in a letter dated 30 June 2024 to the European Commission, which decided not to initiate infringement proceedings against Slovenia. The European Commission explained that, although the Slovenian implementation does not explicitly refer to content that is freely available online as part of the definition of lawful access, it also does not expressly exclude it. Therefore, the CDSM Directive can be considered transposed in accordance with its objective and purpose, as stated in the recitals.

Dr. Maja Bogataj Jančič also addressed the question of who qualifies under ZASP as a beneficiary of the exception for text and data mining for scientific research purposes (Article 57.b ZASP), and she presented the first court case in the EU to deal with this issue: Robert Kneschke v. LAION e.V. (2024). She also pointed out that the second paragraph of Article 57.b ZASP explicitly defines the term “research organisations,” which mainly includes universities (including their libraries) and research institutions.

In her concluding remarks, she stressed the relevance of the Legal Deposit Act (Slo. Zakon o obveznem izvodu publikacij – ZOIPub) as a basis for lawful access, specifically highlighting the role of deposit institutions. According to Articles 13 and 18 of ZOIPub, deposit institutions have a duty – not merely an option – to ensure (lawful) access to deposited publications for researchers and research institutions when used for research purposes. She emphasised that it is essential for deposit institutions to provide only access, and not reproduce the works themselves, as reproduction could trigger additional obligations under ZASP.

She concluded by noting that Slovenia has a very solid copyright framework for building large language models, particularly those covered by the scientific research exception. She also pointed out that in addition to copyright issues, it is necessary to consider legislation on personal data protection (GDPR, ZVOP-2) and the obligations arising from the EU Artificial Intelligence Act (Regulation (EU) 2024/1689 – the so-called AI Act).

Dr. Maja Bogataj Jančič promised all participants that she would publish links to additional resources and literature:

1. PowerPoint presentation “Copyright Foundations for Building Large Language Models in the Slovenian Legal System” is available here:

• PowerPoint file: “Copyright Foundations for Building Large Language Models in the Slovenian Legal System” (Slo. »Avtorsko-pravne podlage grajenja velikih jezikovnih modelov v slovenskem pravnem redu«)

2. The first publication on the new text and data mining (TDM) exceptions is available here:

• Dr. Maja Bogataj Jančič: “Exceptions with teeth: the new Slovenian text and data mining provisions” – contribution for Knowledge Rights 21

3. A paper on the copyright protection of outputs generated by devices was published in the journal of the Institute of Criminology at the Faculty of Law Ljubljana (Law and Artificial Intelligence: Questions of Ethics, Human Rights and Societal Harm, pp. 165–195):

• Dr. Maja Bogataj Jančič: “Can Artificial Intelligence Be an Author of a Copyright Work?” (Slo. »Ali je lahko umetna inteligenca avtor avtorskega dela?«)

4. A paper on text and data mining (TDM) exceptions was published in the Stockholm Intellectual Property Law Review (No. 2/2024, April 2025, pp. 5–8):

• Dr. Maja Bogataj Jančič and Ema Purkart, LL.B.: “Text and Data Mining in the Slovenian Legal System”

5. A short study on the concept of “research organisation” was published by Dr. João Pedro Quintais:

• Dr. João Pedro Quintais: “The concept of ‘research organisation’ and its implications for text and data mining and AI research”

6. A detailed analysis of the Robert Kneschke v. LAION e.V. judgment was published in Pravna Praksa (Vol. 43, Issues 42–43/2024, pp. 12–13) and on the website odipi.si. The full analysis is available here:

• Dr. Maja Bogataj Jančič and Mark Bauer, LL.B.: “Generative AI and Copyright: The First Court Case in Europe”

7. An explanation of why the EU text and data mining (TDM) exceptions (Articles 3 and 4 of the CDSM Directive) can serve as a legal basis for generative artificial intelligence was published on the Knowledge Rights 21 portal:

• Knowledge Rights 21: “Why the EU TDM Exceptions (Art 3 & 4) Can Be Used for AI – Including Generative AI”