Wikimedia and the new knowledge loop – report from a Wikimedia CH roundtable 

by and
Translate this post

One does not need to buy into the hype around generative AI to recognize that it is fundamentally reshaping Wikipedia. The signs are already visible, made clear by two data points from last year: An 8% drop in human traffic, a 50% surge in bot activity. AI is not just intensively reading Wikipedia — it is replacing it as the interface to knowledge. And we do not know how these trends will shape Wikipedia and the wider knowledge ecosystem. What we do know is that Wikipedia is at peak usage and peak risk at the same time.  

In November 2025, Wikimedia CH cooperated with Open Future and IMD Business School and organized a roundtable to answer these questions, and to build an informed response to the rise of generative AI. The event was designed with two key assumptions: that the debate should not be limited to the topic of editing Wikimedia with AI tools; and that we need to invite to the table experts from beyond the Movement: data scientists, AI developers, journalists and researchers.  

The report from this roundtable has been recently published. This is the first step in an initiative, planned by Wikimedia CH for 2026, to define a Wikimedia and AI mission. This mission will include sensemaking and forecasting, starting a technological development roadmap, and actively shaping AI governance. 

The story of two intelligences 

Our inquiry was driven by the concept of collective intelligence, popularized by the French philosopher Pierre Lévy. It is the cognitive capacity that emerges when people collaborate. It is human driven, draws from diverse knowledge sources, and benefits from decentralization and diversity.  

Collective intelligence emerged on Wikipedia 25 years ago, and has been the driving force of the projects, ever since. And using this concept helped us draw out the tension between production of Wikimedia as a knowledge common, and the development of generative AI technologies. The former is based on human collaboration, dialogue and diversity. The latter is often centralized, monolithic and machine based.  

The difference between the two intelligences does not mean that AI cannot be harnessed in support of Wikimedia. And, despite these differences, there is no going back: no plausible scenario where AI influence is avoided. The path forward is one of mixing the two intelligences – while keeping Wikimedia human-centric. 

The challenges of the new knowledge loop   

Web traffic data shows that Wikimedia is being disintermediated: users increasingly obtain from AI services knowledge that they would previously get from Wikipedia. The future threat is not that AI uses Wikipedia, but that Wikipedia becomes invisible to humans. These services are based on Wikimedia content but don’t drive traffic to the platform. If only machines visit Wikipedia, the community model collapses. 

There are signs of a new knowledge loop emerging, in which the AI services will be key actors determining access to knowledge. And the Wikimedia Movement needs to quickly decide what role it wants to play in the ecosystem. Quite possibly, on Wikipedia’s 25th anniversary, we reached the “peak Wikipedia” moment. Continued growth, and ability to pursue the mission of knowledge equity, should not be taken for given.  

Wikimedia needs to play a much more active role in shaping this new ecosystem. Without action, it risks becoming the backbone of AI systems while losing its own sustainability. And in order to do so, it needs to look beyond the boundaries of Wikipedia projects, adopt an ecosystem perspective, taking into account heritage institutions, media outlets, and other commons-based platforms. 

The report from the roundtable looks at these challenges in detail, tackling in turn the paradoxes of openness, the role of Wikipedia as a data commons or differences between public and private knowledge.  

From sensemaking to action 

At the roundtable, one of the participants cut through: why is there so little sense of urgency among Wikimedians? You need to be aware that in several years, the platform might lose its readers, or no longer exist, he told us. You can no longer think of yourself as the disruptor – today, you are being disrupted by AI. And strategic paralysis is as dangerous as giving in to AI hype.  

To address this, Wikimedia Movement needs to lead collective action on behalf of the whole knowledge commons. It is big enough to shape the new ecosystem but can also benefit from the support of many other organizations.  

Learn more

Can you help us translate this article?

In order for this article to reach as many people as possible we would like your help. Can you translate this article to get the message out?