Open Future Foundation

🏖️ Summer Watching or Reading: Common Governance for Open Knowledge Infrastructures On July 14, we hosted an online event highlighting how digital commons transform governance models for critical knowledge infrastructure. Our research fellows, Melissa Hagemann and Dr. Samuel A. Moore, presented papers addressing key questions in Open Access research, knowledge production, and community-led AI governance. If you missed it or want to revisit the discussion, watch the recording and download Melissa's and Samuel's research papers. 📽️ Details here: https://lnkd.in/dUxKEsyN

🔍 Our work on Public AI explores commons-based approaches to machine learning and AI as public infrastructure. We challenge the concentration of power in AI that currently makes diverse activities dependent on a small group of monopolists. At Open Future, we promote alternative public AI systems that demonstrate how more democratic, public interest-oriented technologies can be designed using public AI principles and governance models. A focus of our work is the intersection between AI and openness—specifically addressing the need for established norms around sharing data, compute, models, and code in the emerging field of open AI models. 🕵️ Explore our resource page featuring our published papers, analyses, blog posts, and consultation submissions on this topic. All resources are chronologically organized and easily accessible through our timeline: https://lnkd.in/dkpkVpsf

📌 Save the Date! Join us in this webinar on how data center expansion impacts climate goals—and what we can do about it. 📆 September 16, 2025  ⏰ 16:00-17:00 CEST The rapid growth of AI and digital technologies is fueling unprecedented demand for data centers. This expansion creates serious environmental challenges—threatening climate targets, destabilizing energy systems, and deepening environmental and social inequalities. ECOS is an international NGO advocating for environmentally friendly technical standards, policies, and laws. They provide expertise to decision-makers, promoting strong environmental principles. In this context, we have partnered in a joint report: "How data centre expansion risks derailing climate goals." The webinar will feature insights from co-authors Anastasia Tsougka and Zuzanna Warso. - How can the EU tackle the root causes of digital overconsumption? - What approaches ensure technological progress respects planetary boundaries? Register to find out: https://lnkd.in/dficKvxw

How can CommonsDB improve Wikimedia Commons by connecting digital commons? Wikimedia Sverige is part of the initiative, building a dedicated registry of rights information for public domain and openly licensed images. With Wikimedia Commons now hosting over 100 million media files, managing and improving this vast, freely usable collection is a growing challenge. In a new post, we explore how CommonsDB could help Wikimedians: ✅ Quickly verify copyright information ⬆️ Automate license selection for new uploads 👌 Maintain rights accuracy across millions of files We invite the Wikimedia movement to actively engage in shaping this collaboration and help make CommonsDB as useful and impactful as possible. Read the full article and watch Jan Ainali's introductory video to learn more: https://lnkd.in/dQBvgdUv

Have you seen the official Template for the Public Summary of Training Content for GPAI models released by the EU AI Office at the end of July? Since we've been waiting for and working on this for months, Zuzanna Warso analyzed it to understand where our proposals influenced the template and what it delivers (or omits). One area where the AI Office compromised is transparency in data processing. While we advocated for detailed technical specifications—including filtering processes, anonymization techniques, sampling methodologies, and annotation practices with sufficient depth for meaningful research and auditing—the AI Office instead settled for "general descriptions." This represents a missed opportunity to demonstrate how democratic, public interest-oriented models could be audited based on public AI principles. The consequences are significant: researchers studying gender bias, for example, cannot determine whether training data was filtered for gender-balanced representation, and individuals have no way to verify whether their social media posts were properly anonymized or scraped in their entirety. 🔍 Read the full analysis on our website: https://lnkd.in/d6-dTBUq

🌞 Summer Watching or Reading: What if AI's most powerful tools were in everyone's hands? As we enjoy these calmer summer days, take the opportunity to catch up on our May 20th Public AI event if you missed it—or revisit the discussions about democratizing AI technology. Our white paper, commissioned by Bertelsmann Stiftung through their reframe[Tech] – Algorithms for the Common Good project, explores how state-of-the-art AI can become more accessible and accountable through Public Digital Infrastructure. The research, authored by Felix Sieker, Alek Tarkowski, Cailean Osborne, and Lea Gimpel, offers a promising alternative to private AI dominance: systems built on transparent governance, public accountability, and fair access. The launch event featured compelling insights from industry experts, including Renata Avila's parallel between public education evolution and AI access as a universal right, and Barcelona Supercomputing Center representatives Luca Cominassi and Albert Cañigueral Bagó advocating for fully open-source models. 📚 Download the white paper and watch the recording to learn about: - The systematic Public AI strategy across different layers of the AI stack - The innovative "Gradient of publicness of AI systems" assessment tool 📽️ Details here: https://lnkd.in/dx4gCCHW

Our partners at Wikimedia Sverige introduce #CommonsDB to the Wikimedia Movement in this post published on Diff. The project is building a registry that allows easy checking of whether an image is in the public domain or openly licensed. As they note, "it also offers great potential for the Wikimedia Commons community." 📹 Don't miss Jan Ainali’s video, which clearly and graphically explains how the registry will work using ISCC (International Standard Content Code) and metadata to improve digital content verification. Learn more: https://lnkd.in/ed7VpdAv 

🚀 New Interview Out Now: Decidim and the Spirit of a Digital Commons Community NGI Commons interviewed  Nil Homedes Busquets from Decidim, a groundbreaking platform for participatory democracy. Decidim isn’t just software — it’s a Digital Commons: collectively governed, openly licensed, and globally co-developed. From deep community engagement to sustainability strategies and municipal collaboration, this conversation dives into the practical realities of building democratic digital infrastructure at scale. 👇 Read the full interview with Nil Homedes Busquet and Nicholas Gates: https://lnkd.in/eRFGYccv #DigitalCommons #OpenSource #CivicTech #ParticipatoryDemocracy #PublicInfrastructure #Decidim #NGICommons Martel Innovate OpenForum Europe Open Future Foundation The Linux Foundation CNRS NGI - The Next Generation Internet

🌐 Join us at #Wikimania2025 for an insightful session on "Building a Common Language for the Wikimedia Movement" on August 7 at 13:00 CEST. In this hybrid session, we'll explore how the Wikimedia Movement Advocacy Network is developing a shared vocabulary to effectively communicate about Wikimedia projects as digital commons and digital public goods. The panel features speakers Catalina Carla Frigerio Dattwyler, Bolaji Ayodeji, and Alek Tarkowski and will be moderated by Evelin Heidel. They will dive into the significance of framing Wikimedia projects as digital public goods, highlighting their commons-based governance and role in public digital infrastructure. The session aims to strengthen how Wikimedia's values are communicated to policymakers, editors, and the general public. Join us for this important conversation that builds on Wikipedia's recent recognition as a digital public good. More details here: https://lnkd.in/dv6ZUs7i

The EU AI Office released the official Template for the Summary of GPAI Training Content. While this marks an important step forward, it may not deliver the full accountability it promises. At Open Future, we've been working closely with Mozilla Foundation and experts across sectors to define meaningful AI transparency. Our work has influenced the template's framework, but critical details have been compromised. In her latest analysis, Zuzanna Warso examines where our work made an impact, while identifying concerning compromises such as ambiguous domain disclosure requirements and weak data processing transparency that could allow corporate boilerplate instead of sufficient detail. She also reveals a troubling timeline loophole that gives existing AI systems dominating the market a two-year grace period. 🔍 Read the full analysis on our website: https://lnkd.in/d6-dTBUq