From Data Ownership to Democratic AI
Source
A Critical Q&A on the AI Commons Lab that took place in the AI Commnons Lab, of Mol, Belgium. The document below was produced with the assistance of Dembrane, an AI-driven participation tool for group discussions and focus groups, (see: AI for Deliberation)
URL = https://portal.dembrane.com/en-US/c4561bfe-460c-4e97-8b8c-f3df18c809bb/report
Text
Introduction:
A meeting took place in Mol centered around the idea of an AI Commons Lab, focusing on themes like data ownership, technological sovereignty, and citizen participation. This dynamic Q&A article discusses key insights and dilemmas from reports of the evening, offering critical reflections on the proposals and their societal implications.
1. What is the central goal of the AI Commons Lab in Mol?
The AI Commons Lab in Mol aims to restore citizens' control over their own data and technological environment. As Dembrane explains: "The idea is that individuals become owners of their own data and decide for themselves with whom to share it" [Ailab]. This contrasts with the current model, where large American corporations dominate the internet and associated data.
2. Why is the current digital infrastructure seen as problematic?
The transcripts reveal strong criticism of dependence on American technology and platforms. Dembrane highlights the one-sided data economy: "The technology behind many things we use as a society is American... We need to get off the American drip. That has to change" ([Ailab]). This ties into the broader theme of digital sovereignty and the lack of European alternatives.
3. How is the commons approach concretely applied to technology and data?
The commons model means citizens collectively manage infrastructure and data. Dembrane describes a system where individual data remains in a "personal AI-box" but can also be shared in "local digital twins": "What’s important is that citizens themselves determine what matters... You can use technology like the personal AI-box to share data at a collective, local level" ([Ailab]).
4. What role does open-source software play in this initiative?
Open-source is central because it ensures transparency, ownership, and low barriers to entry. As emphasized: "Thanks to open-source, we can install and use it locally at a very low cost. And all data stays here with us" ([Ailab]). This is seen as essential to breaking reliance on expensive, closed systems from the U.S. or China.
5. Were practical AI applications in a local context discussed?
Yes, several concrete examples were mentioned. These include a local chatbot (trained on municipal website data) that answers residents' specific questions ("When will I get my leaf basket?"). Another is Telraam, where citizens use simple sensors to measure traffic density and optimize mobility. There was also discussion of neighborhood collectives developing smart energy distribution and local data-sharing models.
6. What are the practical challenges in implementing a local data model?
Skepticism exists around feasibility, particularly regarding technological literacy and funding. Beyond technical hurdles ("for many people, it’s complicated"), concerns include costs and accessibility: "GPT is free now, but advanced models are getting more expensive... We want to keep it affordable and ensure data stays local" ([Ailab]).
7. How can AI strengthen citizen participation and democratic decision-making?
A key takeaway is AI’s potential to facilitate public dialogue: "AI can quickly process large amounts of information... That’s citizen participation—getting input from residents to policymakers, and this works really well for that" ([Ailab]). Cities like Genk already use AI-assisted dialogue processes to boost engagement.
8. What role do existing platforms like Dembrane play, and are there risks of new dependencies?
Dembrane, an AI-driven participation tool for group discussions and focus groups, was discussed as an interesting case. It primarily serves governments and municipalities: "We build AI tools [...] to gather input from citizens and stakeholders, enabling faster decision-making at the administrative level."
9. How is the economic model for this infrastructure structured?
Alternative economic models were proposed, such as a token system rewarding contributors of hardware: "Contributions are converted into tokens, and those who help purchase hardware receive a share of token revenues" ([Ailab]). Circular economy principles and local energy ownership (e.g., solar power) were also highlighted.
10. Were there critical remarks or disagreements during the session?
Optimism coexisted with hesitation. While many believe in decentralized data solutions and open-source, concerns linger about privacy, technological complexity, and fragmentation. One critical reflection: "We share everything, but ultimately, everyone must integrate their part into the whole... How do we do that in practice?" ([Ailab]). Some participants stressed the need for experimentation while prioritizing scalability and collaboration.
Conclusion:
The AI Commons Lab in Mol envisions a radically different, citizen-centric digital infrastructure, grounded in data ownership, open-source, and local cooperation. Though the ideals are widely shared, practical, social, and economic challenges—especially around scalability, accessibility, and governance—remain. Yet the examples discussed show that an alternative digital future, closer to citizens, is becoming increasingly tangible."
(https://portal.dembrane.com/en-US/c4561bfe-460c-4e97-8b8c-f3df18c809bb/report)
More information
See for the transcipt of the meeting: