Wikimania 2026: panel on how the far-right affects open knowledge institutions

July 19, 2026

At Wikimania 2026 I participated in this panel, which was held on Wednesday, July 22, 2026. My participation, which involves the project Highlighting the Colombian public domain on Wikidata and Catalan Wikipedia, was shared alongside experiences from the United States, the Netherlands, Germany, Sweden, and Ukraine. Here are the details.

The panel

Wikimania 2026 images

General panel information

Here are the panel details and general information:

The panel video

The session video can be watched at this link, starting at 5:06:26. Here are Larissa's presentation slides.

The case of the systematic attack on Carolina Cárdenas's biography

I presented the case of the systematic attack on Carolina Cárdenas Núñez's biography on Spanish Wikipedia. I also presented this case at Queering Wikimedia. It involves a systematic attack on this biography, where an individual or group deletes any reference to a woman who, 100 years ago, proposed a different model for a woman's role, refusing to conform to that of a submissive and conservative wife, thereby hiding the Colombian artist's potential homosexual relationships. This case stems from the project Highlighting the Colombian public domain on Wikidata, which is also part of the project Strengthening the Latin American Public Domain, developed by Paulina.

The political atmosphere in Colombia is quite polarized. This statement can be understood through the result of the 2026 presidential elections, where the far-right won by a narrow margin of less than 1% of the vote. While Spanish Wikipedia cannot be directly compared to the voting Colombian population, there may be certain parallels that reflect people's sentiment during these times.

What we did to contain these attacks was write the biography in several languages. In addition to having the biography in Spanish, we wrote it in Catalan, which served as a measure to create multiple copies of a type of knowledge that a person or group wants to hide, thereby ensuring its preservation. This translation was also made into Czech and recently into Portuguese. In fact, writing the text in Portuguese uncouples several male figures from the artist's biography, as they implied her achievements were not her own but the result of male influence, taking credit away from her work.

When talking about infrastructures, I forgot to mention that various infrastructures have aspects that offer benefits while simultaneously presenting drawbacks. For example: anonymity on Wikipedia. I defend anonymous editing because it preserves digital rights, but in this case, it works against us because you cannot identify the person or group editing Wikipedia with this conservative, sexist, and homophobic bias. However, at the same time, we believe Wikipedia is one of the few places that allows you to act anonymously, and this in itself is not bad. Moreover, systems to hide IP addresses for anonymous edits were recently expanded.

Another point regarding infrastructure that helps—yet simultaneously impacts—the preservation of Carolina Cárdenas's memory is Wikipedia's policies on content and references. When the person or group editing with conservative, sexist, and homophobic biases does not use references, it amounts to personal opinion. But at the same time, the lack of bibliographic references affects the visibility of women on Wikipedia.

Finally, we can say that the existence of open content does not imply the presence of a feminist point of view. In the Highlighting the Colombian public domain on Wikidata project, we have detected that there is a bias of barely 2 to 4% women in the public domain, while over 90% are men.

Regarding community participation, I proposed that we create and participate in more spaces, such as this panel, events like Queering Wikimedia, and support initiatives like Viquidones UPF, Wikiesfera, and institutions making efforts to make the public domain visible, such as the National Library of Colombia, the Virtual Library of the Banco de la República de Colombia, Wikimedia Colombia, or any others that join in.

My proposal is that we support linguistic diversity, such as using Catalan or other languages to represent our knowledge, and I made a request to everyone participating in the Wikimedia Movement: edit with a feminist point of view, which in Wikipedia jargon would be WP:FPOV.

Presentations at Wikimania 2026

I attended other presentations at Wikimania. Here is a very brief summary of what took place, along with links for more information. The full list of presentations can be viewed on the Wikimania 2026 schedule.

Respecting Indigenous knowledge in open spaces

The title of the presentation is "Keynote: Respecting Indigenous Knowledge in Open Spaces - Indigenous Cultural and Intellectual Property (ICIP) & Indigenous Data Sovereignty (IDSov) across Wikimedia Projects"

Wikimedia Australia created a white paper on culture strengthening guidelines, where the guide serves as a first step toward answering the question: How do Wikimedia platforms recognize Indigenous knowledge? I really liked that each presentation began with greetings in the native languages of the participants. Michelle Collipal, from Chile, spoke about the Mapudungun experience, which is a fundamental part of their identity as Mapuche. They created a protocol regarding who can transmit knowledge and how it should be transmitted, because Mapudungun Wikipedia started in 2006, created by non-Mapuche people. They discussed how to improve free, prior, and informed consent regarding what is represented on Wikipedia.

The New Zealand experience, presented by Migoto Eria, proposed a Māori perspective so that their knowledge is appropriately expressed. She highlighted the colonial perspective by showing us that what is stated in Māori is not the same as what is said in English.

From Canada, Thérèsa Ottawa presented her community's experience. There is an interesting intergenerational effort here.

Resources:

Naming before editing: community autonomy and oral languages in Wikimedia

The description of this session can be viewed at this link. This session focused on how to use Wikimedia projects to keep languages alive. The major challenge here is that orality is paramount for these communities, as speaking is their primary mode of expression. Additionally, societal preconceptions lead to prejudices—such as believing that knowledge in diverse languages is undeveloped or invalid. Wikipedia serves to grant prestige to a language, and what communities ask is that anyone approaching Wikipedia editing commit to learning the language.

I really resonated with what Olga (Bolivia) proposed: contexts differ, so while people in Europe refer to our languages as minority languages, my entire country is not a minority. On the other hand, people feel ashamed of writing poorly in their language, leading to a constant struggle with European academics. How can a European academic dictate to a native speaker how they should speak or write?

To document oral knowledge, recording individuals was not effective. Instead, recording debates proved successful—the exchange of ideas turned out to be far more enriching. There is an ongoing reflection on what communities wish to share or hold back. The main takeaway is that we need to listen deeply. The idea emerged that Wikipedias are created in a specific manner that seems like the only way, even though that isn't true. For instance, the principles of Spanish Wikipedia are complex; one must unlearn them to promote alternative ways of developing diverse Wikipedias in other languages.

One ongoing debate concerns the colonial structure within ISO codes. For example, in Mapudungun, the ISO code used by Wikipedia to categorize the language is ARM, standing for Araucanian. Spanish colonizers used "Araucanians" to refer to the Mapuche people. This acronym categorizes Mapudungun Wikipedia, which is offensive. The same applies to Nahuatl, whose ISO code is PPL (referring to "pipil", meaning young child or child who cannot speak well). This is deeply offensive within the Nahua community. What happens is a repetition of colonial practices—not necessarily directly tied to the Spanish conquest, but rather state-level logics of domination. Wikipedia replicates these ISO practices, but the underlying issue is that ISO relies on colonial methods.

Links:

The Wikimedia Research presentation

The title of this session was "State of Wikimedia Research 2025-2026 + 5 Research Findings and how you can use them in your work. Part 2 : Équité".

Various research studies were presented. The first regarding the Wikipedia gender gap was the study "Monitoring the gender gap in the coverage biology", which demonstrated that when people took the gender gap seriously, campaigns worked, successfully subverting the gender gap from 2017 to 2021.

The second was "How Good is Your Wikipedia? Auditing Data Quality for Low-resource and Multilingual NLP". They audited article quality on Wikipedia and found that small communities conducting human review and writing produce high-value, top-quality edits.

Isaac Johnson delivered a presentation on contributor strategies across Wikimedia projects. He shared 8 key points, including that contributors create task lists, many new editors receive no recognition (inducing uncertainty), mentorship remains a missed opportunity, and there is immense appreciation for human warmth within the Movement.

The presentation "From Wikipedia to AI Answers: Gender Gaps in the Knowledge Pipeline" by Patrick Gildersleve alongside Francesca Tripodi compared search results from AI tools with Wikipedia and analyzed how references are used. They found AI references to be less rigorous. Wikipedia's presence was correlated with high-quality citations. AI performs better when grounded in Wikipedia data. Men receive more coverage from AI tools. Without Wikipedia, citation quality for biographies of women declines significantly in AI tools.

Sabrina Mac Gregor and Nathalie Casemajor presented "Beyond neutrality", discussing notability criteria on French Wikipedia.

The session video is available at this link.

Reference risk: research on source reliability in Wikipedia

The title of this talk is "Reference Risk: A Research-Based Approach to Source Reliability in Wikipedia". Through lists of reliable resources, the number of articles linking to dubious sites decreased. Papers like Longitudinal Assessment of Reference Quality on Wikipedia confirm this finding. Only 14 language editions maintain these curated lists, according to A Comparative Study of Reference Reliability in Multiple Language Editions of Wikipedia.

Resources:

What and who are represented on Spanish Wikipedia? Exploring knowledge representation patterns

I attended this presentation to learn about the methodology used to identify users within the same language project. The methodology involved grouping by country to review the number of Wikidata items and available Wikipedia articles, analyzing everything relative to population size. For instance, Uruguay is small in population, yet has greater representation than Spain, Argentina, or Mexico. They reviewed biographies, and to no one's surprise athlete biographies appeared most frequently. In some cases, athlete biographies exceeded 40% of a country's total biographies. Football (soccer) biographies formed the overwhelming majority, followed by politicians, then artists (both acting and music).

In Argentina, football is a primary focus, as in Uruguay and Colombia. In Mexico and Chile, the main interest lies in politicians. In Spain, there is a diversity of topics. Infrastructure like neighborhoods or transit is a topic of interest in Colombia. Public buildings are a focal point in Chile. In conclusion, the presenter remarked that "territories tell stories."

This turned out to be a replicable methodology, utilizing a WikiProject, queries via the Wikidata Query Service, and exporting results to Google Colab.

Here I asked about analyzing gender gaps in biographies, to which Agustín, the presenter, mentioned it was an area he hadn't delved deeply into yet, but agreed it would be interesting to pursue further.

Links:

The limits of Wikidata

In the session titled "Tabular data vs. structured data: How do we get to a win-win solution?", participants discussed the limits of Wikidata. The thing is, one can upload datasets to Commons, such as in this test I conducted here, which can be displayed in this interactive chart here, featuring data from the Colombian public domain project. The conversation drifted toward Wikidata's capacity limits. Jan mentioned that they hit the first size limit with Q183 (Germany), as overloading the item caused Wikidata to glitch. World population figures break Wikidata because such massive amounts of data cannot be ingested—they mentioned 80 million data points, which I interpreted as annual population logs per country generated by the World Bank.

Tabular data doesn't support time series. Furthermore, when datasets are uploaded to Commons, they cannot be edited directly—you have to re-upload the entire dataset. Back then, the community decided to host datasets on Commons because it made sense at the time, but the approach could be revisited. "How do we prioritize the community wish list?" asked Lydia.

We need solutions. Nevertheless, I appreciated gaining insight into the limits of a tool like Wikidata, as it gives you perspective on how close or far your own project is from reaching community infrastructure limits.

Resources:

Themes of my Wikimania

Apart from the talks, there were other themes I followed, even if some weren't on the official schedule. Conversations over coffee also spark great ideas.

Catalan speakers at Wikimania

A very pleasant surprise was finding more Catalan speakers than expected. There was a Parisian who lived in Barcelona for a few years, with whom I had a conversation in Catalan.

Also, speaking in Catalan with friends from Amical—specifically David Michael Miller, Tiputini, Koldo, Kowalskyn, and Brunaiz, with whom I exchanged greetings in person. I also spoke Catalan with Pablo Aragón and Andreas Kaltenbrunner. It felt good to say "Déu-n'hi-do", a very Catalan expression, now and then.

LED badge messages

These are the messages I used on my LED badge during the Wikimania closing party:

  • Latino pauer
  • e/em/eir
  • What happened on November 6 and 7, 1985?
  • Écoutez la champeta bacana (Q1061165)
  • Pensa en català
  • Non-binary futures
  • Open science, open love
  • Ecofeminismes!
  • Love + kindness = Resistance
  • Crea y mantén un punto de edición

The funding debate

Finally, there are two distinct approaches being debated within the Wikimedia Movement for managing funds. I understand there is a community-driven proposal advocating for specific metrics of achievement to receive funding from the Wikimedia Foundation. On the other hand, a proposal coming from the Foundation suggests distributing funds according to beneficiary achievements—meaning greater outcomes lead to higher funding capacity.

Credits

The images are by VIGNERON, via Commons, and are licensed under Creative Commons Attribution-ShareAlike.