Publications - Pagina 4 van 277

Learning Call [Video]: GenAI and the DSA’s Systemic Risk Framework external link

Quintais, J., Schwemer, S., Çetin, B.S. & Albert, J.A.

2026

Abstract

This Learning Call, part of an ongoing series on current issues surrounding the Digital Services Act (DSA), examines generative AI as a test case for the DSA’s systemic risk framework. Speakers discuss the nature of GenAI‑related risks and the governance mechanisms provided by the DSA and related legal frameworks to address them. Such risks may arise from the dissemination of synthetic and manipulated media, like deepfakes, on very large online platforms and search engines (VLOPSEs), as well as from new GenAI tools deployed by those very platforms in their content moderation systems or via embedded features like AI Overviews. Broadly, the session probes the potential and limits of the DSA’s platform‑centric, risk‑based approach to tackling the complex challenges posed by GenAI, and consider its interaction with other legal instruments (AI regulation, competition law, copyright). The session will also reflect on recent enforcement and policy developments, such as the potential VLOSE designation for services like ChatGPT, the Commission’s investigation into X’s deployment of Grok, the proposed Digital Omnibus on AI, and the review of the EU rules on copyright and AI.

Links

https://dsa-observatory.eu/2026/06/08/genai-and-the-dsas-systemic-risk-framework/

Digital Services Act (DSA), GenAI

RIS

TY  - BLOG
AU  - Quintais, J.
AU  - Schwemer, S.
AU  - Çetin, B.S.
AU  - Albert, J.A.
PY  - 2026
DA  - 2026-06-25
T1  - Learning Call [Video]: GenAI and the DSA’s Systemic Risk Framework
AB  - This Learning Call, part of an ongoing series on current issues surrounding the Digital Services Act (DSA), examines generative AI as a test case for the DSA’s systemic risk framework.

Speakers discuss the nature of GenAI‑related risks and the governance mechanisms provided by the DSA and related legal frameworks to address them. Such risks may arise from the dissemination of synthetic and manipulated media, like deepfakes, on very large online platforms and search engines (VLOPSEs), as well as from new GenAI tools deployed by those very platforms in their content moderation systems or via embedded features like AI Overviews.

Broadly, the session probes the potential and limits of the DSA’s platform‑centric, risk‑based approach to tackling the complex challenges posed by GenAI, and consider its interaction with other legal instruments (AI regulation, competition law, copyright). The session will also reflect on recent enforcement and policy developments, such as the potential VLOSE designation for services like ChatGPT, the Commission’s investigation into X’s deployment of Grok, the proposed Digital Omnibus on AI, and the review of the EU rules on copyright and AI.
K1  - Digital Services Act (DSA)
K1  - GenAI
ER  -

Save .RIS

Bibtex

Online publication{nokey,
	title = {Learning Call [Video]: GenAI and the DSA’s Systemic Risk Framework},
	author = {Quintais, J. and Schwemer, S. and Çetin, B.S. and Albert, J.A.},
	url = {https://dsa-observatory.eu/2026/06/08/genai-and-the-dsas-systemic-risk-framework/},
	year = {2026},
	date = {2026-06-25},
	abstract = {This Learning Call, part of an ongoing series on current issues surrounding the Digital Services Act (DSA), examines generative AI as a test case for the DSA’s systemic risk framework.

Speakers discuss the nature of GenAI‑related risks and the governance mechanisms provided by the DSA and related legal frameworks to address them. Such risks may arise from the dissemination of synthetic and manipulated media, like deepfakes, on very large online platforms and search engines (VLOPSEs), as well as from new GenAI tools deployed by those very platforms in their content moderation systems or via embedded features like AI Overviews.

Broadly, the session probes the potential and limits of the DSA’s platform‑centric, risk‑based approach to tackling the complex challenges posed by GenAI, and consider its interaction with other legal instruments (AI regulation, competition law, copyright). The session will also reflect on recent enforcement and policy developments, such as the potential VLOSE designation for services like ChatGPT, the Commission’s investigation into X’s deployment of Grok, the proposed Digital Omnibus on AI, and the review of the EU rules on copyright and AI.},
	keywords = {Digital Services Act (DSA), GenAI},
}

Save .bib

Citizens’ attitudes toward AI in journalism and authenticity: Evidence from six countries on four different continents external link

Morosoli, S., Mattis, N., Naudts, L., Vreese, C.H. de & Helberger, N.

Journal of Quantitative Description: Digital Media, vol. 6, pp: 1-53, 2026

Abstract

Artificial intelligence (AI) is reshaping journalism at an unprecedented pace, challenging long-held assumptions about what makes news credible, authentic, and human. As AI systems increasingly generate, curate, and personalize information, societies around the world are grappling with questions about how this transformation affects trust in the media and the perceived integrity of journalism. Scholars have highlighted that AI not only changes newsroom practices but also redefines audiences' expectations of journalistic values such as accuracy and transparency. Yet, public perceptions of these changes remain under-explored from a global, comparative perspective, as previous research has mainly taken a Western perspective. This study, therefore, examines to what extent individuals' attitudes toward AI in journalism and perceptions of authenticity differ across six culturally diverse countries: Brazil, Denmark, Japan, the Netherlands, South Africa, and the United States. Using nationally representative surveys (N = 6,961), we analyze four core dimensions: (1) journalistic principles for AI-supported news, (2) individual engagement intentions for AI-supported news, (3) general authenticity perceptions across different dimensions and (4) the perceived impact of AI on them. Our findings show significant cross-national heterogeneity in attitudes toward AI in journalism, with more pronounced optimism in South Africa and Brazil and more pronounced skepticism in Western countries. Moreover, we frequently find ambivalent attitudes, as well as high degrees of heterogeneity within individual countries as well, with self-reported generative AI use emerging as one explanatory factor. Engagement intentions demonstrate a trend for ignoring AI-supported news, particularly in Western contexts, alongside a universal reluctance to share such content. This reflects broader trust issues and reputational concerns. Despite these differences, respondents across all countries converge on the importance of the human element when it comes to authenticity, rooted in journalists' character and values.

Links

Artificial intelligence, Journalism

RIS

TY  - JOUR
AU  - Morosoli, S.
AU  - Mattis, N.
AU  - Naudts, L.
AU  - Vreese, C.H. de
AU  - Helberger, N.
PY  - 2026
DA  - 2026-09-15
T1  - Citizens’ attitudes toward AI in journalism and authenticity: Evidence from six countries on four different continents
JO  - Journal of Quantitative Description: Digital Media
VL  - 6
SP  - 1-53
DO  - https://doi.org/10.51685/jqd.2026.019
AB  - Artificial intelligence (AI) is reshaping journalism at an unprecedented pace, challenging long-held assumptions about what makes news credible, authentic, and human. As AI systems increasingly generate, curate, and personalize information, societies around the world are grappling with questions about how this transformation affects trust in the media and the perceived integrity of journalism. Scholars have highlighted that AI not only changes newsroom practices but also redefines audiences\' expectations of journalistic values such as accuracy and transparency. Yet, public perceptions of these changes remain under-explored from a global, comparative perspective, as previous research has mainly taken a Western perspective. This study, therefore, examines to what extent individuals\' attitudes toward AI in journalism and perceptions of authenticity differ across six culturally diverse countries: Brazil, Denmark, Japan, the Netherlands, South Africa, and the United States. Using nationally representative surveys (N = 6,961), we analyze four core dimensions: (1) journalistic principles for AI-supported news, (2) individual engagement intentions for AI-supported news, (3) general authenticity perceptions across different dimensions and (4) the perceived impact of AI on them. Our findings show significant cross-national heterogeneity in attitudes toward AI in journalism, with more pronounced optimism in South Africa and Brazil and more pronounced skepticism in Western countries. Moreover, we frequently find ambivalent attitudes, as well as high degrees of heterogeneity within individual countries as well, with self-reported generative AI use emerging as one explanatory factor. Engagement intentions demonstrate a trend for ignoring AI-supported news, particularly in Western contexts, alongside a universal reluctance to share such content. This reflects broader trust issues and reputational concerns. Despite these differences, respondents across all countries converge on the importance of the human element when it comes to authenticity, rooted in journalists\' character and values.
K1  - Artificial intelligence
K1  - Journalism
ER  -

Save .RIS

Bibtex

Article{nokey,
	title = {Citizens’ attitudes toward AI in journalism and authenticity: Evidence from six countries on four different continents},
	author = {Morosoli, S. and Mattis, N. and Naudts, L. and Vreese, C.H. de and Helberger, N.},
	url = {https://journalqd.org/article/view/9452},
	doi = {https://doi.org/10.51685/jqd.2026.019},
	year = {2026},
	date = {2026-09-15},
	journal = {Journal of Quantitative Description: Digital Media},
	volume = {6},
	pages = {1-53},
	abstract = {Artificial intelligence (AI) is reshaping journalism at an unprecedented pace, challenging long-held assumptions about what makes news credible, authentic, and human. As AI systems increasingly generate, curate, and personalize information, societies around the world are grappling with questions about how this transformation affects trust in the media and the perceived integrity of journalism. Scholars have highlighted that AI not only changes newsroom practices but also redefines audiences\' expectations of journalistic values such as accuracy and transparency. Yet, public perceptions of these changes remain under-explored from a global, comparative perspective, as previous research has mainly taken a Western perspective. This study, therefore, examines to what extent individuals\' attitudes toward AI in journalism and perceptions of authenticity differ across six culturally diverse countries: Brazil, Denmark, Japan, the Netherlands, South Africa, and the United States. Using nationally representative surveys (N = 6,961), we analyze four core dimensions: (1) journalistic principles for AI-supported news, (2) individual engagement intentions for AI-supported news, (3) general authenticity perceptions across different dimensions and (4) the perceived impact of AI on them. Our findings show significant cross-national heterogeneity in attitudes toward AI in journalism, with more pronounced optimism in South Africa and Brazil and more pronounced skepticism in Western countries. Moreover, we frequently find ambivalent attitudes, as well as high degrees of heterogeneity within individual countries as well, with self-reported generative AI use emerging as one explanatory factor. Engagement intentions demonstrate a trend for ignoring AI-supported news, particularly in Western contexts, alongside a universal reluctance to share such content. This reflects broader trust issues and reputational concerns. Despite these differences, respondents across all countries converge on the importance of the human element when it comes to authenticity, rooted in journalists\' character and values.},
	keywords = {Artificial intelligence, Journalism},
}

Save .bib

How does political efficacy condition clicks on politics? Understanding information-selection behavior in algorithmic feeds over time

Wan, J., Araujo, T., Helberger, N. & Vreese, C.H. de

Journal of Computer-Mediated Communication, vol. 31, iss. : 5, 2026

Abstract

This study uses political efficacy to explain individual differences in selecting political information on algorithmic feeds. Combining survey data and 186 Dutch individuals’ usage histories of Google Discover, a fully algorithmically curated feed, we examined both within-person changes of information-selection behavior over time and between-person differences conditioned by political efficacy in three aspects: the proportion of political information selected by users, exposure diversity as manifested via sources, and proportion of negative political information. While we did not find any longitudinal change conditioned by external efficacy on information-selection patterns, we found that internally efficacious users are more resilient to prevalent concerns about algorithms’ negative influence on political information consumption and exposure diversity. We observed their stable proportions of political information consumption and consumption from increasingly diverse sources over time. Alarmingly, there is a growing divide in exposure diversity between less and more efficacious users over time. Altogether, our study not only highlights users’ agency in navigating algorithmic environments but also underscores the importance of empowering users with account of different vulnerabilities.

Links

DOI: https://doi.org/10.1093/jcmc/zmag022

RIS

TY  - JOUR
AU  - Wan, J.
AU  - Araujo, T.
AU  - Helberger, N.
AU  - Vreese, C.H. de
PY  - 2026
DA  - 2026-09-17
T1  - How does political efficacy condition clicks on politics? Understanding information-selection behavior in algorithmic feeds over time
JO  - Journal of Computer-Mediated Communication
VL  - 31
IS  - 5
DO  - https://doi.org/10.1093/jcmc/zmag022
AB  - This study uses political efficacy to explain individual differences in selecting political information on algorithmic feeds. Combining survey data and 186 Dutch individuals’ usage histories of Google Discover, a fully algorithmically curated feed, we examined both within-person changes of information-selection behavior over time and between-person differences conditioned by political efficacy in three aspects: the proportion of political information selected by users, exposure diversity as manifested via sources, and proportion of negative political information. While we did not find any longitudinal change conditioned by external efficacy on information-selection patterns, we found that internally efficacious users are more resilient to prevalent concerns about algorithms’ negative influence on political information consumption and exposure diversity. We observed their stable proportions of political information consumption and consumption from increasingly diverse sources over time. Alarmingly, there is a growing divide in exposure diversity between less and more efficacious users over time. Altogether, our study not only highlights users’ agency in navigating algorithmic environments but also underscores the importance of empowering users with account of different vulnerabilities.
ER  -

Save .RIS

Bibtex

Article{nokey,
	title = {How does political efficacy condition clicks on politics? Understanding information-selection behavior in algorithmic feeds over time},
	author = {Wan, J. and Araujo, T. and Helberger, N. and Vreese, C.H. de},
	doi = {https://doi.org/10.1093/jcmc/zmag022},
	year = {2026},
	date = {2026-09-17},
	journal = {Journal of Computer-Mediated Communication},
	volume = {31},
	issue = {5},
	abstract = {This study uses political efficacy to explain individual differences in selecting political information on algorithmic feeds. Combining survey data and 186 Dutch individuals’ usage histories of Google Discover, a fully algorithmically curated feed, we examined both within-person changes of information-selection behavior over time and between-person differences conditioned by political efficacy in three aspects: the proportion of political information selected by users, exposure diversity as manifested via sources, and proportion of negative political information. While we did not find any longitudinal change conditioned by external efficacy on information-selection patterns, we found that internally efficacious users are more resilient to prevalent concerns about algorithms’ negative influence on political information consumption and exposure diversity. We observed their stable proportions of political information consumption and consumption from increasingly diverse sources over time. Alarmingly, there is a growing divide in exposure diversity between less and more efficacious users over time. Altogether, our study not only highlights users’ agency in navigating algorithmic environments but also underscores the importance of empowering users with account of different vulnerabilities.},
}

Save .bib

Why Human Rights Lawyers Should Care About Intellectual Property external link

Izyumenko, E. & Geiger, C.

Human Rights Here, 2026

Links

https://www.humanrightshere.com/post/why-human-rights-lawyers-should-care-about-intellectual-property

RIS

Save .RIS

Bibtex

Save .bib

From Upcycling to Book Destruction: Copyright Exhaustion’s Environmental Paradox external link

Izyumenko, E.

Kluwer Copyright Blog, 2026

Links

https://legalblogs.wolterskluwer.com/copyright-blog/from-upcycling-to-book-destruction-copyright-exhaustions-environmental-paradox/

Copyright, upcycling

RIS

Save .RIS

Bibtex

Save .bib

Governing the Post-Quantum Cryptography Transition in the US and the EU: A Comparative Analysis external link

Jančiūtė, L. & van Hoboken, J.

2026

Abstract

Large scale quantum computers are anticipated to pose serious threat to the current digital infrastructure in that they will enable decryption of much of information encrypted with conventional methods used today. This threat has been gaining momentum as a next cybersecurity frontier. Post-quantum cryptography (PQC) is deemed to be the main solution in mitigating this threat and the necessity to act upon it has already been embraced by numerous governments around the world. This article analyses how two jurisdictions-the US and the EU-have been governing their transition to PQC. It discusses the state of play, the relevant actors, policies as well as the main challenges and future trajectories. This process has been unfolding in different ways in the US and the EU which are now both pursuing the same ultimate quantum-readiness timeline of 2035. Alongside, the considerations surrounding PQC algorithmic choices are also explored.

Links

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7364718

quantum technologies

RIS

TY  - UNPD
AU  - Jančiūtė, L.
AU  - van Hoboken, J.
PY  - 2026
DA  - 2026-08-29
T1  - Governing the Post-Quantum Cryptography Transition in the US and the EU: A Comparative Analysis
AB  - Large scale quantum computers are anticipated to pose serious threat to the current digital infrastructure in that they will enable decryption of much of information encrypted with conventional methods used today. This threat has been gaining momentum as a next cybersecurity frontier. Post-quantum cryptography (PQC) is deemed to be the main solution in mitigating this threat and the necessity to act upon it has already been embraced by numerous governments around the world. This article analyses how two jurisdictions-the US and the EU-have been governing their transition to PQC. It discusses the state of play, the relevant actors, policies as well as the main challenges and future trajectories. This process has been unfolding in different ways in the US and the EU which are now both pursuing the same ultimate quantum-readiness timeline of 2035. Alongside, the considerations surrounding PQC algorithmic choices are also explored.
K1  - quantum technologies
ER  -

Save .RIS

Bibtex

Working paper{nokey,
	title = {Governing the Post-Quantum Cryptography Transition in the US and the EU: A Comparative Analysis},
	author = {Jančiūtė, L. and van Hoboken, J.},
	url = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7364718},
	year = {2026},
	date = {2026-08-29},
	abstract = {Large scale quantum computers are anticipated to pose serious threat to the current digital infrastructure in that they will enable decryption of much of information encrypted with conventional methods used today. This threat has been gaining momentum as a next cybersecurity frontier. Post-quantum cryptography (PQC) is deemed to be the main solution in mitigating this threat and the necessity to act upon it has already been embraced by numerous governments around the world. This article analyses how two jurisdictions-the US and the EU-have been governing their transition to PQC. It discusses the state of play, the relevant actors, policies as well as the main challenges and future trajectories. This process has been unfolding in different ways in the US and the EU which are now both pursuing the same ultimate quantum-readiness timeline of 2035. Alongside, the considerations surrounding PQC algorithmic choices are also explored.},
	keywords = {quantum technologies},
}

Save .bib

Het naburig deepfakerecht voorbij: strikt behoud zeggenschap, met eerlijke vergoedingen voor beperkte exploitatiebehoefte? : Empirische inzichten uit de Nederlandse muzieksector download

Valk, E.G.

Auteursrecht, iss. : 3, pp: 91-100,

Abstract

Toenemende kansen en risico’s van deepfaketechnologie hebben geleid tot nabuurrechtelijke oplossingen in onder meer Nederland en Denemarken. De (initiatief)wetsvoorstellen uit beide landen worden in dit artikel vergeleken. De Nederlandse benadering met een naburig ‘deepfakerecht’ voor eenieder is in het bijzonder onvolkomen, mede gezien het toereikende bestaande wettelijk kader. De analyse in dit artikel toont op basis van resultaten van een recente survey in de Nederlandse muzieksector aan dat de naburige rechtenbenadering voor enkel uitvoerend kunstenaars evenmin past. Vervolgens worden alternatieve oplossingen uitgewerkt voor de juridische uitdagingen (zoals wegtekenrisico’s) die in de creatieve sector spelen met betrekking tot toestemming en eerlijke vergoedingen voor deepfakegebruik.

Links

Auteursrecht_2026_3

Auteursrecht, deepfakes, exploitatierechten, wet op de naburige rechten

RIS

TY  - JOUR
AU  - Valk, E.G.
PY  - 
DA  - DATE ERROR: pub_date = 
T1  - Het naburig deepfakerecht voorbij: strikt behoud zeggenschap, met eerlijke vergoedingen voor beperkte exploitatiebehoefte? : Empirische inzichten uit de Nederlandse muzieksector
JO  - Auteursrecht
IS  - 3
SP  - 91-100
AB  - Toenemende kansen en risico’s van deepfaketechnologie hebben geleid tot nabuurrechtelijke oplossingen in onder meer Nederland en Denemarken. De (initiatief)wetsvoorstellen uit beide landen worden in dit artikel vergeleken. De Nederlandse benadering met een naburig ‘deepfakerecht’ voor eenieder is in het bijzonder onvolkomen, mede gezien het toereikende bestaande wettelijk kader. De analyse in dit artikel toont op basis van resultaten van een recente survey in de Nederlandse muzieksector aan dat de naburige rechtenbenadering voor enkel uitvoerend kunstenaars evenmin past. Vervolgens worden alternatieve oplossingen uitgewerkt voor de juridische uitdagingen (zoals wegtekenrisico’s) die in de creatieve sector spelen met betrekking tot toestemming en eerlijke vergoedingen voor deepfakegebruik.
K1  - Auteursrecht
K1  - deepfakes
K1  - exploitatierechten
K1  - wet op de naburige rechten
ER  -

Save .RIS

Bibtex

Article{nokey,
	title = {Het naburig deepfakerecht voorbij: strikt behoud zeggenschap, met eerlijke vergoedingen voor beperkte exploitatiebehoefte? : Empirische inzichten uit de Nederlandse muzieksector},
	author = {Valk, E.G.},
	url = {https://www.ivir.nl/nl/publications/het-naburig-deepfakerecht-voorbij-strikt-behoud-zeggenschap-met-eerlijke-vergoedingen-voor-beperkte-exploitatiebehoefte-empirische-inzichten-uit-de-nederlandse-muzieksector/auteursrecht_2026_3/},
	year = {},
	date = {DATE ERROR: pub_date = },
	journal = {Auteursrecht},
	issue = {3},
	pages = {91-100},
	abstract = {Toenemende kansen en risico’s van deepfaketechnologie hebben geleid tot nabuurrechtelijke oplossingen in onder meer Nederland en Denemarken. De (initiatief)wetsvoorstellen uit beide landen worden in dit artikel vergeleken. De Nederlandse benadering met een naburig ‘deepfakerecht’ voor eenieder is in het bijzonder onvolkomen, mede gezien het toereikende bestaande wettelijk kader. De analyse in dit artikel toont op basis van resultaten van een recente survey in de Nederlandse muzieksector aan dat de naburige rechtenbenadering voor enkel uitvoerend kunstenaars evenmin past. Vervolgens worden alternatieve oplossingen uitgewerkt voor de juridische uitdagingen (zoals wegtekenrisico’s) die in de creatieve sector spelen met betrekking tot toestemming en eerlijke vergoedingen voor deepfakegebruik.},
	keywords = {Auteursrecht, deepfakes, exploitatierechten, wet op de naburige rechten},
}

Save .bib

Podcast The Binary Agora: The AI Act’s Watermarking Rules external link

Quintais, J.

The Binary Agora, iss. : 48, 2026

Abstract

Podcast with João Pedro Quintais about the European Commission's Guidelines on transparency obligations for providers and deployers of certain AI systems and the application of Article 50 of the EU AI Act more broadly. João also commented on how copyright considerations and history can make us aware of challenges that may arise in the context of the AI Act.

Links

RIS

Save .RIS

Bibtex

Save .bib

Journalism as Data: GDPR Implications of Licensing Journalistic Content to Large Language Models external link

Bouchè, G. & Steketee, M.

Technology and Regulation, pp: 102-121, 2026

Abstract

This article explores the data protection implications under the GDPR of integrating journalistic content into Large Language Models (LLMs). The number of commercial partnerships between AI companies and news publishers for the licensing of daily news and archival content has rapidly increased. We contend that while LLMs and their top-layer applications do offer innovative solutions for news dissemination, publishers should carefully evaluate their position under the GDPR. Comparing different technical solutions available, in particular pre-training, fine-tuning and Retrieval Augmented Generation (RAG), we analyse the relevant regulatory barriers and opportunities, focusing in particular on the distribution of processing roles, lawfulness and transparency of these deals, and the application of the special regime for journalistic processing under art 85(2) GDPR.

Links

GDPR, Journalism, language models, Privacy

RIS

TY  - JOUR
AU  - Bouchè, G.
AU  - Steketee, M.
PY  - 2026
DA  - 2026-08-27
T1  - Journalism as Data: GDPR Implications of Licensing Journalistic Content to Large Language Models
JO  - Technology and Regulation
SP  - 102-121
DO  - https://doi.org/10.71265/7yhpdx81
AB  - This  article  explores  the  data  protection  implications  under  the  GDPR  of  integrating  journalistic  content  into  Large  Language  Models  (LLMs).  The  number  of  commercial  partnerships between AI companies and news publishers for the licensing of daily news and  archival  content  has  rapidly  increased.  We  contend  that  while  LLMs  and  their  top-layer applications do offer innovative solutions for news dissemination, publishers should carefully evaluate their position under the GDPR. Comparing different technical solutions available,  in  particular  pre-training,  fine-tuning  and  Retrieval  Augmented  Generation  (RAG), we analyse the relevant regulatory barriers and opportunities, focusing in particular on  the  distribution  of  processing  roles,  lawfulness  and  transparency  of  these  deals,  and  the application of the special regime for journalistic processing under art 85(2) GDPR.
K1  - GDPR
K1  - Journalism
K1  - language models
K1  - Privacy
ER  -

Save .RIS

Bibtex

Article{nokey,
	title = {Journalism as Data: GDPR Implications of Licensing Journalistic Content to Large Language Models},
	author = {Bouchè, G. and Steketee, M.},
	url = {https://techreg.org/article/view/25844},
	doi = {https://doi.org/10.71265/7yhpdx81},
	year = {2026},
	date = {2026-08-27},
	journal = {Technology and Regulation},
	pages = {102-121},
	abstract = {This  article  explores  the  data  protection  implications  under  the  GDPR  of  integrating  journalistic  content  into  Large  Language  Models  (LLMs).  The  number  of  commercial  partnerships between AI companies and news publishers for the licensing of daily news and  archival  content  has  rapidly  increased.  We  contend  that  while  LLMs  and  their  top-layer applications do offer innovative solutions for news dissemination, publishers should carefully evaluate their position under the GDPR. Comparing different technical solutions available,  in  particular  pre-training,  fine-tuning  and  Retrieval  Augmented  Generation  (RAG), we analyse the relevant regulatory barriers and opportunities, focusing in particular on  the  distribution  of  processing  roles,  lawfulness  and  transparency  of  these  deals,  and  the application of the special regime for journalistic processing under art 85(2) GDPR.},
	keywords = {GDPR, Journalism, language models, Privacy},
}

Save .bib

Alternative Compensation Systems

Quintais, J.

In: Elgar Encyclopedia of Intellectual Property Law, , Edward Elgar Publishing, 2025, ISBN: 9781800886926

Links

DOI: https://doi.org/10.4337/9781800886926.alternative.compensation.systems

alternative compensation systems, collective rights management, Copyright

RIS

Save .RIS

Bibtex

Save .bib

Top Keywords

Abstract

Links

RIS

Bibtex

Abstract

Links

RIS

Bibtex

Abstract

Links

RIS

Bibtex

Links

RIS

Bibtex

Links

RIS

Bibtex

Abstract

Links

RIS

Bibtex

Abstract

Links

RIS

Bibtex

Abstract

Links

RIS

Bibtex

Abstract

Links

RIS

Bibtex

Links

RIS

Bibtex