A shift in how knowledge is shared – OKFN Newsletter Oct/Nov 2026
Publicado em: A shift in how knowledge is shared – OKFN Newsletter Oct/Nov 2026
Hello Open Knowledge community,
Sharing is hard to sustain when the incentive is extraction. This tension lies at the heart of The Sharing Shift, whose new website we are launching this month. Since 2024, we have collaborated with the Centre for Internet and Society (CIS at CNRS) to develop the legal, social, and technical infrastructure necessary to maintain the shareability of knowledge in the era of AI. This includes licences, governance models, standards, code, social practices, and strategic litigation. In our new white paper, written with Renata Avila, Ramya Chandrasekhar, Melanie Dulong de Rosnay and Andrew Rens, we envision the governance of Digital Public Infrastructures as Commons.
The same question runs through our AI Learning Labs. The Brazil and Uruguay pilots travelled beyond their original scope: the World Economic Forum used one to show why provenance matters, and in Barcelona Open Tech Week we ran a live query against Uruguay’s national energy data. In Abrelatam/Condatos Guatemala, we walked archivists through digitising collections for a world of AI agents. And the Climate Academy Chatbot (CABot)’s team turned its lessons into a free course on School of Data, ‘How to Build a Responsible AI Chatbot Using Open Technologies’.
Two conversations worth your time this month. As CKAN turns 20, we spoke with OKFN Board Member Irina Bolychevsky about how it became the default for national open data portals, now behind 1,000+ catalogues in 60+ countries. And documentation is now online from The PeaceTech We Want Summit, held with the PeaceTech Alliance on 19 October: a practitioner-led conversation, grounded in a peacebuilder-first, Do No Harm approach, on rooting peacebuilders in the design of AI and peacetech tools rather than building for them from outside.
Much of this work now lands in rooms where standards are decided. We are in Paris for the DPGA Annual Members Meeting, making the case that attribution and provenance belong in the DPG Standard, and at the Paris Peace Forum asking what the G7’s Vision on AI Openness means in practice. This month, our CEO, Renata Ávila, helped close the UN CSTD Working Group on Data Governance. Today, she is in Brussels for NGI’s Digital Commons Policy Summit, discussing how the new Digital Commons European Digital Infrastructure Consortium can work with the commoners who maintain it.
There is more: Open Knowledge Austria is active again in Vienna, with Open Data Inside as a first step. We are consulting on Open Data Day 2027, and you can find experts (and nominate yourself) through our Global Directory.
Enjoy the reading.
Introducing The Sharing Shift
This week we are launching the official website for The Sharing Shift, our global effort – underway since 2024 – to keep data and knowledge shareable, and collectively valuable, in the age of AI. With the Centre for Internet and Society (CIS at CNRS), we study, develop and test open licences, governance models, standards, metadata and code, alongside social practices and strategic litigation. Four focus areas guide the work, from creating a public-interest ecosystem to shifting who benefits from sharing.
White paper: ‘Envisioning Digital Public Infrastructures as Commons’
Download The Sharing Shift’s latest release: a white paper by Renata Avila, Ramya Chandrasekhar, Melanie Dulong de Rosnay and Andrew Rens. ↪ Read it here
Event: ‘Shaping Openness for Digital Public Goods in the Age of AI’
The Digital Public Goods Alliance (DPGA) and Linagora, together with The Sharing Shift, are convening a closed-door workshop on November 16, 2026 to discuss AI systems as Digital Public Goods and as digital commons. ↪ Learn more
Shaping the Open Transition: Inside the OTR Symposium 2026
On 26–27 October, the first Open Technology Research Symposium took place at the Universitat de Barcelona under the theme ‘Shaping the Open Transition’. Across 22 sessions in six thematic blocks, researchers and practitioners examined how the shift to open technologies actually happens — and how that research can inform the policy decisions shaping public institutions’ use of technology. The symposium closed with OpenForum Europe‘s Basil Cousins Award, presented to Yigakpoa S. for her paper on governing digital commons as civic infrastructure. The Open Source Initiative then hosted an AI Salon on openness in AI and its implications for global policy.
The PeaceTech We Want: Summit Documentation in Video
With forty countries facing armed conflict, technology is shaping peace and war in ways its creators rarely intend. On 19 October, together with the PeaceTech Alliance, we held The PeaceTech We Want Summit — an online, practitioner-led conversation grounded in a peacebuilder-first, Do No Harm approach. It brought peacebuilders, technologists, researchers and policymakers together to ask how we root peacebuilders in the design of AI and peacetech tools, rather than building for the outside. The full framing and speaker line-up, including Lena Slachmuijlder, Brett Solomon, Milos Strugar and Miriam Ferrer-Coronel, are now online.
↪ Check out the full documentation
How to Build a Responsible AI Chatbot: A New Course on School of Data
CABot began as an experiment: a climate chatbot grounded in trusted literature rather than the open web. Now the #semanticClimate team has turned those lessons into a free course on School of Data. Over five modules and 35 lessons, ‘How to Build a Responsible AI Chatbot Using Open Technologies’ takes you from how large language models generate language (and why they hallucinate) through retrieval-augmented generation to a working chatbot architecture. Responsibility runs throughout: reducing hallucinations, citing sources, and weighing fairness, privacy and sustainability. Everything is open source, with the full repository on GitHub for you to adapt.
Coming soon: a course version in Hindi, Bengali, Tamil and Telugu. Enroll now in English and we’ll let you know when they launch.
From Guatemala City to Barcelona: Taking Responsible AI on the Road
Our AI Learning Labs travelled this month – first to Abrelatam/Condatos in Guatemala City, then to MozFest during Open Tech Week in Barcelona. In Guatemala, we ran two sessions along different tracks. With AVANCSO’s Alejandro Flores-Aguilar and Sara Fratti of ILDA, Andrés Vázquez (OKFN) walked through five steps for digitising physical archives facing AI — from assessing collections to choosing formats and metadata that keep records usable and open. With Brazil’s CGU and Uruguay’s AGESIC, we shared lessons from the MCP pilot connecting AI to open data portals, including what broke and what we’d do differently. In Barcelona, Kannika Thaimai and Patricio Del Boca demonstrated a live query against Uruguay’s National Energy Balance, showing how a Model Context Protocol can bridge AI tools and trusted open data. You can try the pilot yourself.
‘CKAN Has Spread Because It Makes Maintaining Data More Cost-effective and Stimulates the Economy Through Data Reuse’
For the 22nd conversation in our 100+ Conversation series, we spoke with digital strategist and OKFN board member Irina Bolychevsky — CKAN’s first non-developer hire, in 2011. As CKAN turns 20, she traces how a small side project became the default open data platform, now powering more than 1,000 portals in over 60 countries. She explains the decisions behind that scale: narrowing the pitch to governments, investing in CKAN 2.0, and building a community where each government contract funded features everyone could use. Her caution for the AI era: data is contextual, and infrastructure should not be led by AI’s interpretation of it.
Why Traceability Matters: Our Pilots in the World Economic Forum
The World Economic Forum has published a piece on how organizations can share knowledge that AI can use responsibly — and it draws directly on our AI Learning Labs pilots with Brazil and Uruguay. In one example from Uruguay’s public energy data, the AI attributed a change partly to climate factors, but the dataset contained no climate data to support that explanation. The lesson is not that AI fails, but that provenance makes failure visible and correctable. The article also notes how partners’ frequently asked questions informed testing and helped refine guidance, echoing what we’ve learned: clear definitions, documented limitations, and example questions are as important as the model itself.
Attribution, Commons and Open Source AI: Our Sessions at the DPGA Annual Members Meeting
This month we are in Paris for the Digital Public Goods Alliance Annual Members Meeting, hosted by the Government of France. The AMM brings the DPGA community together for three days of learning, exchange and collaboration. We will be there throughout, with our sessions concentrated on 18 November. In Enabling Attribution for AI Systems through DPGA Standard Indicators, alongside Creative Commons, Open Future, OSI and the DPGA, we’ll share lessons from our MCP pilot with Brazil and Uruguay on moving beyond attribution towards provenance and trust. Shielding the Commons, with the Wikimedia Foundation, is an interactive workshop testing a policy framework that treats open knowledge systems as critical digital infrastructure – participants will work through realistic national crisis scenarios. And with OSI, Creative Commons and the DPGA Standards Team, we’ll get into the practical barriers to implementing the Open Source AI Definition.
Final Session of the UN Data Governance Working Group
On 4–5 November, the UN CSTD Multi-Stakeholder Working Group on Data Governance held its eighth and final session in Geneva, reviewing the chapters of its Progress Report and agreeing to convey it to the 30th session of the CSTD in April 2027, before it reaches the 81st UN General Assembly. The report covers fundamental principles, benefit-sharing, interoperability and safe cross-border data flows. Our CEO, Renata Ávila, sits on the Working Group as a non-State member and co-facilitatedTrack 2 on interoperability between national, regional and international data systems as well as contributed as drafter of the report, the thread that runs through much of our own work on open data infrastructure.
OKFN at the Paris Peace Forum
Thinking about AI in the long term means asking how we build resilience in our economies, for our environment, and for future generations. That was the frame for ‘Building Resilient AI Futures through Openness and Cooperation’, a thematic event of the 9th Paris Peace Forum under the G7 France Presidency, held at Inria with GPAI, hosted at the OECD-OCDE. Nearly 100 people came together to ask what the G7’s Vision on AI Openness, adopted in May, means in practice: open how, for whom, and to what purpose? Renata Ávila and Ramya Chandrasekhar joined discussions on access versus agency, the strain on shared infrastructure, language inclusion, and why openness can strengthen security — if verification keeps pace.
Today in Brussels: OKFN at NGI’s Digital Commons Policy Summit
Today, our CEO Renata Ávila joins Session 3 of the Digital Commons Policy Summit 2026 in Brussels. Moderated by Valérian Guillier of CNRS, the session DC-EDIC and the Commoners: Operationalisation, Cooperation and Ambition asks how the newly established Digital Commons European Digital Infrastructure Consortium moves from structure to practice, and how it can work in genuine partnership with the commoners who build and maintain shared digital infrastructure. It is the practical question behind much of our own work on digital public infrastructure, and one we are glad to see at the centre of the European agenda.
Tracing Governmental Data at the Media Party Buenos Aires
Media Party Buenos Aires 2026 gathered journalists, developers and media innovators under the theme ‘Storytelling in the Age of Agents’. Among the lightning talks was our senior software developer Andrés Vázquez, who shared what we learned from connecting AI to open data through a Model Context Protocol. Drawing on the pilots with the governments of Uruguay and Brazil, he showed how MCP lets people ask questions in natural language and follow the answer back to the dataset, a practical route from open data to trustworthy, traceable AI answers. The full programme and recordings are online.
Twenty-five years ago, Nature and Science published the first draft of the human genome. To mark it, the University of Barcelona invited OKFN Board Member Tim Hubbard, one of the authors of those landmark papers, to open its anniversary programme with a lecture on the past, present and future of genomics. He traced the economic and scientific challenges behind the project, made the case for researchers having access to data within secure environments, and pointed to the possibilities AI now offers for genomic research — with optimism for the future about sequencing’s use in treating patients. A recording of the lecture is now available.
Congratulations to ROOST Coop, whose inclusion on the Digital Public Goods Alliance registry we’re glad to see. ROOST builds open source safety infrastructure for detecting harmful content and putting guardrails around AI systems. That openness matters most where safety is concerned. When detection and moderation capabilities are open, they can be audited, localised for different languages and contexts, and deployed by smaller organisations that could never build them alone.
Check out the latest highlights from our global movement.
Open Knowledge Austria Starts Anew with Focus on AI
An OKFN chapter in Austria is now reactivated. A new Austrian organisation for open data, shared knowledge and technology that serves the public good, it began in Vienna and carries forward a story that first took shape in 2011, when the Open Knowledge Forum Austria began supporting municipal open data portals, open science and data literacy. The former chapter dissolved at the end of 2018; today’s OKAT is a new organisation taking up that cause. One concrete starting point is the relaunch of Open Data Inside: a way to make the people and projects using open data more visible.
Reclaim the Open! Nordic–Baltic meetup of open knowledge communities
Reclaim the Open explores openness as a core Nordic–Baltic value and considers knowledge and culture as critical infrastructure in times of geopolitical and technological disruption. Together, we will map ongoing initiatives, share experiences, and spark lasting collaboration across the Nordic and Baltic region.
OK North invites people working on or interested in these questions across culture, research, media, technology, and public policy.
OK North is a Nordic–Baltic network connecting people and organisations working with open knowledge, digital commons, and public policy. The network brings together Open Knowledge Finland, Open Knowledge Sweden, Open Knowledge Estonia, Creative Commons Finland, Creative Commons Sweden, Creative Commons Norway, and OpenDenmark, and is supported by the Nordic Council of Ministers.
We Want to Hear From You Before Start Planning Open Data Day 2027
We are consulting our communities on what the central theme of Open Data Day in 2027 should be. What are the key priorities for those working directly with open data in government and organisations? In a digital world increasingly enclosured by AI, what kind of future do we want to build for open data? Please help us by completing the survey.
Discover Experts from Around the Globe on Open Knowledge’s Directory
Who are the French-speaking civic tech experts I could ask for advice on my project?
Our Global Directory is designed precisely to help you sort this type of question. It is a curated list of real people with direct contact details, a valuable resource for collaboration and coordination in the open movement worldwide. You can discover experts from around the globe, grouped into 43 categories related to openness, and in 30+ languages. You can also nominate experts, including yourself.
These are the latest discussions on the Open Knowledge Forum. Have your say!
@jqnatividad on Standards & Interoperability ↗
“The Greece National Data Portal shows a quality score next to their data. Wouldn’t it be great if more data portals do this – perhaps, by using a reusable quality standard and/or CKAN extension?”
@mackeynichols on Sites ↗
“Toronto Open Data Portal is the November Site of the Month! City of Toronto Open Data staff here! Thrilled we’re featured.”
@jze on Frictionless Data ↗
“Just now I realized once again how useful it would be if frictionless-py could also read Shape files. Is there anyone else interested in that topic?”
Save the dates!
Internet Governance Forum 2026
Nairobi
14–18 December
The Internet Governance Forum (IGF) is a global multistakeholder platform that facilitates discussions on public policy issues pertaining to Internet governance. The IGF was one of the most important outcomes of the United Nations World Summit on the Information Society (WSIS) that mandated the United Nations Secretary-General to formally convene the Forum.
Geneva AI Summit 2027
Switzerland
21—22 June, 2027
The Geneva AI Summit 2027 unites two strategic priorities: AI as a driver of prosperity and progress for all and fostering trustworthy, responsible and safe use of AI. Innovation and trust reinforce each other, responsible AI requires transparency and broad accessibility. Together, these priorities position Switzerland as both a global AI leader and a constructive bridge-builder in shaping AI’s future.
About us.
Open knowledge is any content, information or data that people are free to use, re-use and redistribute — without any legal, technological or social restriction.
The Open Knowledge Foundation’s mission is to create a fair, sustainable and open digital future, advancing open knowledge as a design principle beyond just data. We do it by guiding and supporting the creation of digital infrastructure, developing policies and methodologies, harnessing communities and advocating for literacies and standards in a sustainable, ethical and agile manner.
Feed: Open Knowledge Blog
Url: blog.okfn.org