On September 22, I had the honor of serving as a keynote panelist at the Open Access Scholarly Publishing Association (OASPA) Conference. The panel addressed a crucial set of questions under the title “Who owns open knowledge? Who has the power and responsibility?”—issues at the very heart of scholarly communication and open science. My contribution to the discussion was an impulse text titled “Who Owns, Produces, and Certifies Open Knowledge?”, in which I explored these themes from both an information science and a sociological perspective. I am very grateful to Joachim Schöpfel, with whom I was able to discuss the content a few days ago, for his valuable comments.
In the following, I am publishing the full text of my contribution.
What is (open, scholarly) knowledge?
In everyday use, “knowledge” often just means facts or rules—like knowing when a store closes. In academia, though, knowledge is the outcome of inquiry: traceable, testable, and methodologically sound. Open knowledge, as defined by the Open Definition, is knowledge made legally and technically free—to access, use, modify, and share without restriction.
Who can make use of (open) knowledge?
While it formally belongs to everyone, ownership isn’t the central issue—the real question is who can use and shape it. What really matters is who can make effective use of open knowledge—and that depends according to Pierre Bourdieu on access to various forms of capital:
- Social capital: networks and relationships
- Cultural capital: prior knowledge or credentials
- Financial capital
- Symbolic capital: recognition, status, and prestige
Without these, open knowledge remains technically available—but not equally usable.
Production and certification
A deeper issue is emerging—not just due to APC-funded models, but also growing political influence:
Does all valuable or truthful knowledge really have an equal chance to be published—and then made open? Who is even in a position to produce (open) knowledge?
This raises three key follow-up questions on the level of certification:
- Who gets to certify open knowledge as trustworthy?
- Are they qualified to do so? Peer review and database rankings are often debated.
- Do they apply standards fairly in practice? There are signals that the standards are not always perfectly applied: Failures include poor-quality publications slipping through or dubious journals being indexed.
Ultimately, it’s less about who owns open knowledge and more about:
- Who can produce it?
- Who can certify it?
Also the change and innovation are primarily driven by producers, who can initiate it, and certifiers, who can legitimize it.
Political penetration and scholarly autonomy
Even in traditionally open societies, open-access services are being defunded, and researchers from certain regimes (identified by IP addresses) face restricted access to open knowledge. More concerning, however, is growing political influence over what gets published, shifting control away from the scientific community.
Censorship can mean two things: blocking access to knowledge (as it is the case when access is prevented via excluding IP addresses), or preventing its production. While the former can regularly be bypassed, the latter is more serious—and we’re now seeing cases where certain kinds of knowledge are not meant to be produced at all.
Who can protect the freedom to produce knowledge?
Democratic states that respect academic autonomy are essential—without that, knowledge becomes politically compromised. But then, how should certifiers like reviewers and databases respond?
If peer review works as intended, state-influenced, low-quality research should not be published. The responsibility lies with the scientific community to uphold standards and resist political pressure. But like courts or central banks, scientific communities can be compromised. Both the power to produce and the responsibility to certify knowledge are vulnerable to political influence.
Ideally, states and scientists would share these roles. But in reality—with authoritarian states restricting openness and even democratic ones steering research agendas—we must ask: Where can science still be safely and independently practiced?
This situation is oddly reminiscent of 2016, when the Internet Archive scrambled—but succeeded—in setting up mirrors to prevent a politically motivated shutdown or erasing critical information from the archive. It’s no coincidence that the Technische Informationsbibliothek Hannover in Germany (TIB) in Hanover is now building a dark archive for arXiv.
Open Knowledge and AI
Of course I also want to touch AI: According to the data-information-knowledge-wisdom (DIKW) pyramid of the International Society for Knowledge Organization (ISKO) , we move from data at the base, to information, then knowledge, and finally wisdom at the top.
https://www.isko.org/cyclo/dikw
AI, however—at least when its outputs aren’t curated—tends to remain at the lower levels: it replaces knowledge with data and information.
This distinction matters, especially as AI systems increasingly interact with open content. For instance: is blocking AI bots from indexing open repositories—due to server load—a violation of open knowledge principles of restriction-free access/ use?
And if we want to strengthen author ownership, we’d need to ensure proper attribution—even in AI-driven contexts. But since AI can process open content freely, that’s hard to enforce.
One potential step: a voluntary commitment from publishers to only allow access to AI systems that honor attribution. Besides the fact that excluding these machines from indexing open content is technically hard to implement, this raises another dilemma: would that content still count as truly open? Such measures could require technical restrictions that undermine the very ethos of open knowledge.
How to preserve the integrity of open knowledge?
We were asked to propose a concrete way to preserve the integrity of open knowledge—here is mine.
I support the development of community-driven frameworks—such as a Trusted Open Knowledge Label—that help define and promote responsible openness.
This means going beyond access and focusing on key values like:
- Attribution, especially in the context of AI systems that reuse and remix content at scale;
- Transparency in review and certification, so that trust in open content is earned and not assumed;
- And an explicit awareness of the structural barriers—social, political, and economic—that shape not only who produces knowledge, but whose knowledge is recognized and legitimized.