by Meghan O’Neill (Librarian and Open Access Repository Officer, IMEC)

The Belgian OA Networking Event, held at Hoek 38 in Brussels on 2 June, brought together members of the Belgian scholarly communication community to exchange experiences and discuss two timely topics: Diamond Open Access and the impact of AI scraping on repositories. Beyond exploring these themes, the event aimed to strengthen the Belgian OA Network through practical discussion and knowledge sharing. 

The morning started off with a deep dive into Diamond: pragmatic presentations by experts Johan Rooryck (Leiden University) and Jonathan Dumont (ULiège). Johan described his hands-on experience flipping Glossa: A Journal of General Linguistics to a Diamond OA publishing model. He reflected on the importance of sovereignty in Diamond OA publishing and called for equal excellence and quality across Diamond OA journals. Most importantly, he celebrated diamond as an egalitarian publishing model wherein multilingual journals can thrive, and members of the community have an opportunity to take on tasks traditionally reserved for publishers. Jonathan Dumont presented on a few exciting and interconnected national projects. For example, the strategy and workflow behind the fully Diamond OA journal Circular Materials and Chemistry – A Cross-Disciplinary Journal (CMC) born out of the FWB-CoARA initiative. Jonathan emphasised the importance of creating Diamond OA journals as tangible proof that this publishing model can succeed, particularly in STEM disciplines where such initiatives are still relatively scarce. The presentation wrapped up with guidelines on Diamond OA business models and an introduction to The BeDiamond Initiative: a national project to establish a foundation to promote Diamond publishing within Belgium and support diamond journals. 

Panelists Demmy Verbeke (KU Leuven), Johan Rooryck, Jonathan Dumont and Inge Van Nieuwerburgh (UGent) took part in a lively exchange that began with acknowledging the buzz term of the morning: digital sovereignty. In fact, this concept of sovereignty is not only relevant for the community to maintain control and curation of Diamond OA journals but for institutional repositories, and the entire scholarly communications network. The hour-long panel discussion grappled with facets of Diamond OA policy such as the importance of scholarly governance at international and regional levels, recognised by UNESCO (the UNESCO Recommendation on Open Science from 2021). It was clear that national diversity and collaboration between Flanders and Wallonia is key for Belgium to continue to carve out a place within the European context.  Practical initiatives were also discussed, such as training for librarians and research support staff, redirecting APC expenditure towards sustainable publish-and-read initiatives, and placing greater emphasis on research quality rather than journal impact metrics. 

Specialist Maurane Raskinet (ULiège) started the afternoon sessions with a succinct overview of the impact of AI scraping on repositories. Maurane introduced one of the recurring dilemmas of the afternoon: distinguishing between “good bots” and “bad bots.” According to a 2025 COAR survey, 90% of repositories reported aggressive AI bot activity. Legal expert Thomas Margoni (KU Leuven) shed light on many facets of national copyright laws in the EU applicable to libraries and cultural heritage institutions in the context of AI with reference to Articles 3 and 4 of the Copyright in the Digital Single Market (CDSM) Directive. He urged that the role of licenses should be rethought. For example, he investigated the role of CC licenses as a social and ethical agreement placed under the umbrella of overarching copyright law. The final expert speaker of the afternoon continued the conversation of rapidly advancing technology. Christina Vandendyck (KBR) gave a detailed introduction to BelgicaWeb a BRAIN 2.0 project funded by BELSPO which aims to make the born-digital cultural heritage of Belgium FAIR (findable, accessible, interoperablee and reusable). She detailed many technical attributes of the project such as software decisions and demonstrated the crawling mechanism which uses Browsertrix crawler software to gather content from the web and social media.  
 
Rather than aiming for definitive answers, the afternoon sessions encouraged participants to exchange practical experiences, identify common challenges and reflect upon realistic approaches within their own institutions. We divided into groups facilitated by volunteer organizers and supported by experts to ensure opportunities for engagement. The legal group benefited from the additional expertise of Leona King (KU Leuven) and Francesca Portante d’Alessandro (KU Leuven), while the technical group was supported by experts Patrick Hochstenbach (UGent) and Luc Bastiaensen (UAntwerp). Although it is difficult to summarize such rich conversations, each group discussion contributed to the key takeaways and action items which wrapped up the event. We discussed the criticality of clearly defining the role of repositories for repository managers and including this role into repository policies. For example, what does the concept of trust look like in your repository, and how would that be included in your policy? Considering the increase of poorly behaved bot attacks, it is also crucial to understand and clearly define what behaviour is “bad” and why. How can we decide what bots we let in? How can we protect our repositories from those that are not welcome?  

The bot debate is intrinsically linked with legal issues and goes beyond technical solutions. As Thomas Margoni discussed, distinguishing between the good and bad governance of the internet should not be up to a select few commercial organizations (read: Cloudflare). In fact, the constant “are you human?” question asked by such software places an additional burden on researchers and on those seeking information from a user-friendliness point of view. Raising awareness of these issues during trainings for researchers will be useful, without making it too overwhelming. 
 
The legal discussions highlighted that licenses alone will not solve the challenges posed by AI scraping. Participants questioned how licenses can be enforced in practice, particularly when repository content is reused globally. This underlined the need for coordinated national and international policy discussions beyond the repository level. 
 
The complexity of technical and legal issues at play necessitates repository managers, librarians, research support, and all members of the scholarly communication ecosystem to stay connected, to be informed about the latest developments in COAR and IETF, and to proceed with collective action. Beyond scholarly communication specialists, collaboration with institutional leadership, policy and legal experts is needed to navigate the way forward for repositories.

Rather than providing definitive answers, the Belgian OA Networking Event demonstrated the value of bringing together repository managers, librarians, legal experts and policy makers to exchange experiences and identify practical ways forward. As AI technologies continue to evolve, maintaining this dialogue within the Belgian Open Access Network will remain essential.