Open Future Foundation

☃️ Holiday Season Alert: The Open Future team will be taking a well-deserved break from 24 December to 5 January. A huge thank you to our committed team members, donors, partners, and the entire Digital Commons and Open Internet community. Your dedication and support have made 2024 truly enlightening. Here's to an exciting next year! ✨ #DigitalCommons #OpenInternet

🔍 Check out this research article by Dr. Kacper Szkalej and Prof. Dr. Martin Senftleben, published at Jipitec. It covers the application of Share-Alike obligations (in Creative Commons licences) to trained models, curated datasets, and AI output by studying how #GenerativeAI training intersects with openly licensed works. 🎯 https://lnkd.in/dvchzKPt A previous version of the study was published on our website in June. The piece, commissioned to the Institute for Information Law (IViR), showed the impact of SA and #copyleft licensing on machine learning and generative AI. The clauses of these licenses provide a mechanism that many commons-based projects rely on to prevent missappropiation of open knowledge resources, large-scale collaborative projects and #DigitalCommons in general. 🎯 https://lnkd.in/dR_9VP6N

🤫 Do trade secrets really stand in the way of meaningful AI training data transparency? Find out why such claims are misleading in this new analysis by Maximilian Gahntz & Zuzanna Warso.  https://lnkd.in/dsKfrvMF #EUAIAct #AITransparency #AIEthics

ℹ️ The "sufficiently detailed summary" that the EU AI Act requires developers of General-Purpose AI (GPAI) models to publish offers the most significant opportunity yet to advance transparency around training data. This analysis by Zuzanna Warso (Director of Research at Open Future) and Maximilian Gahntz (AI Policy Lead at the Mozilla Foundation) argues what information should be disclosed in the summary, who it matters to, and why trade secrets cannot be used as an excuse for a lack of transparency. The EU's AI Office will provide a template for the "sufficiently detailed summary" of training data before next summer. 🤔 Read the analysis and join the conversation: https://lnkd.in/dsKfrvMF #AIAct #AITransparency

🔉 Organized by ProductNation/iSPIRT, the DPI & AI Indo-French Workshop will provide recommendations for the AI ACTION Summit from December 11 to 13. A discussion panel about Digital Commons and Public Digital Infrastructure will take place this Thursday from 11:25 to 12:10 CET. What is the role of the state, the market and communities in Digital Public Infrastructures? Jan Krewer (Senior Policy Analyst, Open Future), Astha Kapoor (Co-Founder and Director, Aapti Institute), Mila Samdub (CyberBRICS Project, Fundação Getulio Vargas; Open Future's fellow) and Renata Avila (CEO, Open Knowledge Foundation) will talk about this and other questions in a session moderated by Ramya Chandrasekhar (PhD Researcher, Centre Internet et Société (CIS) of CNRS) with Hari Subramanian (iSPIRT Foundation) and Didier Calcei (Associate professor, ISG Institut Supérieur de Gestion). 🎟️ Registration for online participation is still open: https://lnkd.in/esiT_zgg #DigitalCommons #PublicDigitalInfrastructure

🗞️ "Transparency around the content used to train AI and information about how it was processed can support the legitimate interests of preventing discrimination and respecting cultural diversity." Learn more in this article by Maximilian Gahntz & Zuzanna Warso, published in Tech Policy Press. https://lnkd.in/dz9FU-Sa #AIAct #AICommons

🤖 Open Future and Fundacja Centrum Cyfrowe’s latest research provides ideas on how to support #opensourceAI and #publicAI globally, learning from cases of small language model development in Poland: SpeakLeash | Spichlerz and PLLuM (Polish Large Language Model) initiative. The publication is written by Alek Tarkowski (Open Future), Kuba Piwowar (Fundacja Centrum Cyfrowe) and Michał Owczarek (Uniwersytet SWPS). This report highlights how #digitalcommons can compete with commercial models and help fight concentrations of power in AI. 🔗 Read or download: https://lnkd.in/dE3XAV8e

🏗️ What is Digital Public Infrastructure (DPI)? This opinion from our fellow Mila Samdub and Chand Rajendra-Nicolucci seeks to bring more specificity and clarity to the concept. “Multiple visions of DPI are circulating today: DPI-as-roads and DPI-as-radio, as well as other variations that don’t map quite so readily onto a single metaphor, including an ongoing European push for DPI as systems to support regional sovereignty.” https://lnkd.in/e-YPVDpu #DPI #DigitalPublicInfrastructure #DigitalCommons

📚 Check out The Glushko & Samuelson Information Law and Policy Lab and Open Future report "Demonopolizing the European Public Domain" for insights on Google Books exclusivity clauses and the Open Data Directive. Written by Arilee Arends, Annette de Bont and Myrthe Rosenberg under the supervision of João Pedro Quintais and Ot van Dalen, the publication provides actionable recommendations for librarians, policymakers, and digital activists. 🔍 Read or download here: https://lnkd.in/dsfe8ZuS #ResearchReport #OpenData #DigitalCommons

💌 Did you miss the monthly newsletter we published last Friday? Learn about our latest reports on #PublicAI and #PublicDomain, get details on November workshops and discussions on #DigitalCommons, register for upcoming events, and more. Read the issue and sign up for the next one ⬇️ https://lnkd.in/dG9_bC5k