<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by The Data Tank on Medium]]></title>
        <description><![CDATA[Stories by The Data Tank on Medium]]></description>
        <link>https://medium.com/@TheDataTank?source=rss-14056a7befdf------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*oB3kcv4lzJF0-6yBJQHiTQ.png</url>
            <title>Stories by The Data Tank on Medium</title>
            <link>https://medium.com/@TheDataTank?source=rss-14056a7befdf------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 18:43:28 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@TheDataTank/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com" rel="hub"/>
        <item>
            <title><![CDATA[Critical Next Steps for EU Policy: Strengthening Media’s Bargaining Power over Generative AI]]></title>
            <link>https://medium.com/@TheDataTank/critical-next-steps-for-eu-policy-strengthening-medias-bargaining-power-over-generative-ai-1feba9b6933e?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/1feba9b6933e</guid>
            <category><![CDATA[democracy]]></category>
            <category><![CDATA[media]]></category>
            <category><![CDATA[copyright]]></category>
            <category><![CDATA[fair-remuneration]]></category>
            <category><![CDATA[genarative-ai]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Wed, 04 Nov 2026 16:47:08 GMT</pubDate>
            <atom:updated>2026-11-06T14:55:24.983Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Anna Colom</strong>, The Data Tank | <strong>Paul Keller</strong>, Open Future | <strong>Mira Milosevic</strong>, Global Forum for Media Development (GFMD)| <strong>Marta Poblet, </strong>The Data Tank | <strong>Wout van Wijk, </strong>News Media Europe</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vTzpnNLJbUl9j4J-vqEz_w.png" /><figcaption>‘The AI-Deal’ by Daniela Zampieri/ <a href="https://betterimagesofai.org">https://betterimagesofai.org</a> / <a href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</a></figcaption></figure><p>We are at a critical juncture for media, information integrity and democracy. Major structural changes that began with social media in the 2010s are now accelerating, driven by a few powerful tech companies dominating General Purpose AI (GPAI) models. This power is compounded by extractive market and operational practices and the way Large Language Models and their AI assistants and agents unreliably but confidently regurgitate information while seeking to monopolise the attention of the audience. This, at the costs of public interest and independent media.</p><p>To address these challenges, technical solutions tackling content attribution, preventing unsolicited scraping, building digital skills, and setting newsroom guidelines remain essential at the newsroom level. However, the <a href="https://cdn.prod.website-files.com/63bd953ecb882b8e559c1614/6a0b38fae86b009d3027b07e_LitReview%20Media%20and%20GenAI%20DataTank%20May18.pdf">literature is clear </a>on the need for ‘holistic solutions that strengthen the collective and bargaining power of media’ involving ‘a multiplicity of instruments, including effective regulatory approaches and the creation of and investment in alternative public digital infrastructure’.</p><p>Policy ambitions do not automatically translate into increased bargaining power, effective enforcement, or viable alternatives. Instead, targeted initiatives are needed that tackle the current power asymmetries straight on.</p><p>The window for action is open but time is running out. So, drawing on the proposals shared by <a href="https://luma.com/824xze49">seven leading experts in this field</a> during the panel discussion ‘Strengthening Media’s Bargaining Power over Generative AI: Next Steps for EU Policy’, –some also co-authors of this article– we summarise the proposed critical next steps as follows:</p><p><strong>On fair compensation and copyright: roll out regulation fit for the current age, rather than expanding inadequate policies.</strong></p><p>Firstly, where works are used to train and operate generative AI systems, the current context requires a shift from a framework built on exclusive rights to one built on remuneration. This concerns the use of works as inputs for generative AI, not copyright in general. Insisting on licensing under exclusive rights reproduces the asymmetry it is meant to fix: large AI providers choose whom to pay, and smaller and minority-language media are left out. A statutory remuneration right, attached to the commercial deployment of generative AI systems rather than only to training, would ensure that all media whose work underpins these systems share in the value they create, without depending on individual bargaining power. As long as licensing continues alongside such a remuneration right, it needs guardrails. Paul Keller, Director of Policy at Open Future, proposes prohibiting exclusive content deals between media organisations and very large platforms or gatekeepers. Exclusive deals turn content into a competitive moat: they entrench the position of the largest AI providers, shut out smaller and European developers, and leave media dependent on a single buyer.</p><p>However, as reflected in the current draft of <a href="https://unesdoc.unesco.org/ark:/48223/pf0000398418">UNESCO’s guidance on fair compensation for news</a> there is no one pathway that will work alone or everywhere and the areas of regulation cut across: (i) Rights based arrangements, (ii) Competition-based arrangements, and (iii) Fiscal or levy based arrangements.</p><p>Second, there is a need for new enforcement tools. Current AI systems involve content and copyright uses such as training, retrieval, summarisation, and others that do not map into the traditional relationship between a publisher and a search or aggregation platform.</p><blockquote>Digital platforms cannot simply assume that the value created by press publishers is available to them to use for free. But, big players can at the moment choose whom to negotiate and whether to negotiate to the detriment of particularly smaller publishers.</blockquote><p>There is therefore a need for tools for enforcement, such as investigative powers and remedies that can be imposed, complemented by measures including meaningful transparency mechanisms, a shift in the burden of proof, and a scalable approach to licensing. Licensing schemes alone do not solve the fact that AI summaries now include answers so audiences move away from publishers, leading to disintermediation. The media sector needs to be more pragmatic, unite and ask for viable market alternatives and these involve not only licensing but also competition tools. Competition or media authorities can impose or intervene to restore fair competition, reduce or eliminate bottlenecks, mitigate dependencies and provide alternatives.</p><p><strong>Competition policy and digital regulation to address structural dependencies, decentralise power, and create the conditions for diverse, thriving alternative ecosystem.</strong></p><p>As stated during the event by Maria Luisa Stasi, Head of Law and Policy for Digital Markets at Article 19, in addition to a new approach to licensing, and tools for enforcement and its monitoring, there is a need for alternatives to the current technology and infrastructure themselves in order to dilute the power of and dependence on large AI corporations. This goal can be achieved by opening up the market and controlling gatekeeping power, supporting new and diverse entrants inclusive of small players and public interest technologists, acting at the level of incentives but also financial support, and using tools such as R&amp;D and procurement. These measures need to be accompanied by an enabling ecosystem based on portability, interoperability and enhanced transparency.</p><p>Content licensing marketplaces, platforms that let publishers license their work to AI companies without having to negotiate a bilateral deal with each one, are another piece of infrastructure taking shape, largely built by some of the same companies whose systems created the problem in the first place. As <a href="https://www.brookings.edu/articles/same-gatekeepers-new-tollbooths-in-the-ai-content-licensing-market/">argued by Courtney C. Radsch</a>, if the sector’s whole strategy is to negotiate a better position inside marketplaces owned and governed by Big Tech, we risk ending up dependent on infrastructure that cannibalises public interest media and its content. A strategy that works for media and democracy should not be relying solely on ‘their’ systems. In Europe, the direction of travel is either consolidation into larger infrastructure companies like Cloudflare, or collective licensing led by collective management organisations such as the <a href="https://dpcmo.dk/">Danish Press Publications’ Collective Management Organization (DPCMO)</a>. Few independent marketplaces are accessible to small and community media, especially those in minority languages with no strong collecting society behind them.</p><p><strong>Centering the public interest: the importance of principles and normative frameworks</strong></p><p>As argued during the event by Prof. Natali Helberger, co-Founder of the AI, Media and Democracy Lab, there is a need for a new knowledge and innovation ecosystem that serves public interest alternatives. Asking who is in the best position to develop public interest alternatives, and how these parties can be enabled and incentivised is an important first step into the right direction. Next to this we need more agreement of what genuinely public-interest oriented infrastructures mean, also from the perspective of citizens, and what values should guide their development.</p><p>Underpinning these decisive regulatory measures, then, are values and principles, as argued by Saorla McCabe, Deputy Secretary to the IPDC — UNESCO:<strong> ‘</strong>It is important that any policy effort in this area is built on a common vision, shared values and normative foundations’ including ’human rights, freedom of expression, and media pluralism’. McCabe, who co-led the Development of UNESCO’s ‘Guidance on Fair Compensation for News’, also reminded us during the event that these policy efforts need to take into account that these challenges affect the very foundations of democratic societies.</p><p><strong>Horizontal empowerment and collective action</strong></p><p>Gaining bargaining power entails being bolder and owning the infrastructure itself, empowering existing collective initiatives and proceeding with creating its own marketplaces as a media sector.</p><p>Organisations and newsrooms need support to join initiatives like SPUR, the publishing-industry coalition behind the Content Telemetry Standard, which shows a publisher whether their content has been crawled, cited or used to ground an AI answer. Likewise, they can consider collectives such as the common data spaces, and notably the European media data space, or follow the examples of collectives, such as Mozilla Data Collective, one of the few examples of a non-profit platform governed by its contributors, built on community-sourced data.</p><p>Ultimately, enforcing regulation and addressing these power asymmetries is essential and urgent, not only to address the sustainability of the media sector and the integrity of information, but also to protect democratic resilience and the diversity of languages and high quality information providers of different kind and contexts. Policy on copyright and fair compensation needs to favour sector-wide remuneration over individually negotiated licensing under exclusive rights, which mainly benefits large players. Policies aiming at addressing the current power asymmetries and sustaining a thriving and diverse media sector need to be complemented by competition measures and enforcement, and support an alternative knowledge and infrastructure ecosystem. This ecosystem needs to be grounded in public interests principles, as well as on principles of transparency, portability and interoperability, and offer a meaningful alternative to a handful of concentrated technology players. This creates new opportunities, also for smaller players.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1feba9b6933e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Taking the Temperature on Data Access: Do We Need a Data Barometer?]]></title>
            <link>https://medium.com/@TheDataTank/taking-the-temperature-on-data-access-do-we-need-a-data-barometer-b38c48e23e27?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/b38c48e23e27</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-stewardship]]></category>
            <category><![CDATA[data-access]]></category>
            <category><![CDATA[open-data]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Mon, 05 Oct 2026 08:50:27 GMT</pubDate>
            <atom:updated>2026-10-05T08:50:27.768Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Stefaan G. Verhulst</strong>, The Data Tank | <strong>Silvana Fumega</strong>, Global Data Barometer | <strong>Paulina Behluli</strong>, The Data Tank</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yRSeJGc3u5WlRU63dgalcg.png" /></figure><p>At this year’s <a href="https://dataforpolicy.org/">Data for Policy Conference</a> in Barcelona, we tried something new: a live vote to take the temperature of the room. The question: is access to data moving in the right direction?</p><p>The room included data practitioners, policymakers, researchers, industry representatives and civil society: all working on different ends of the same data pipeline. Each sees a different part of the picture, and we wanted to see what happens when those views are put side by side.</p><h3>A mixed reading</h3><p>The result was mixed, and honestly, that feels right. There was no clear verdict that things are getting better or worse — although the majority (34 vs 24) felt that we are experiencing a data winter. Some respondents blamed this “winter” on AI data extraction, a lack of legal tools for access, shrinking social media data access, and declining open data.</p><p>As Stefaan Verhulst wrote last year, borrowing from Dickens, it is <a href="https://sverhulst.medium.com/it-was-the-best-of-times-it-was-the-worst-of-times-the-dual-realities-of-data-access-in-the-age-8732837acaa5">the best of times and the worst of times for data access</a>. On one side, privacy-enhancing technologies, synthetic data, data commons and generative AI itself could open a <a href="https://arxiv.org/abs/2405.04333">Fourth Wave of Open Data.</a> On the other, platforms have closed research access, climate data is being privatised, open government data programmes are stalling, certain governments are questioning whether public data should be charged, and “AI anxiety” makes data holders reluctant to share.</p><p>Left unchecked, those pressures point toward a <a href="https://sverhulst.medium.com/are-we-entering-a-data-winter-f654eb8e8663">data winter</a>: a period in which data that could serve the common good is frozen and immobilised. Depending on where you sit, you can look at the same ecosystem and see either the spring of hope or the winter of despair.</p><h3>The one thing they agreed on</h3><p>The sample was small, so the numbers should be read with care. But one point drew near-universal agreement: without a barometer, it is very hard to tell whether the data ecosystem is moving forward, and in which direction.</p><p>We track weather, inflation, unemployment and air quality with regular, shared measures. For the broader data ecosystem, the closest thing we have is the <a href="https://globaldatabarometer.org/">Global Data Barometer</a>. Its second edition, released in May 2025, assesses 43 countries, with a focus on Latin America and the Caribbean and Africa. It scores them on data governance, capabilities and availability, with AI readiness, inclusion and data use as cross-cutting themes.</p><p>Its findings echo the mixed mood in our room. Many countries have adopted data governance laws, but practice lags well behind policy. Former open data champions are struggling to sustain their programmes. Inclusion remains largely symbolic, and AI is moving faster than the rules meant to govern it.</p><p>The Global Data Barometer gives us a rigorous, country-level baseline. What our vote added was a quick pulse from the people working in the system. The two are complementary: one measures the structures, the other captures how those structures feel on the ground.</p><h3>An experiment, not a finished tool</h3><p>This was a first attempt to combine a visual, interactive format with research on the state of data access. It worked well enough to start a conversation, and it showed what needs more design: clearer indicators aligned with the Global Data Barometer, explicit questions on <a href="https://datastewards.net/news">data stewardship</a>, a larger and more varied sample, and a way to repeat the reading over time.</p><p>You can explore the <a href="https://data-summer-winter-show.lovable.app/">prototype barometer here</a>. We would welcome ideas on what it should measure and who should take part. Taking the temperature is only a starting point. A next step is to better understand what may sit behind these perceptions: whether responsibilities for enabling responsible data use are clear, whether organisations are able to collaborate across sectors, and whether data can be accessed, shared and reused responsibly. These are all dimensions of data stewardship, and we are interested in exploring how they could be assessed comparatively across countries. If we want the data ecosystem to move in the right direction, we need a way to know which direction it is moving, and to better understand the conditions that enable data to be accessed and used for the public good.</p><p>If you are interested to help us shape this experiment better, reach out to us at: <a href="mailto:silvana@globaldatabarometer.org">silvana@globaldatabarometer.org</a> &amp; <a href="mailto:contact@datastewards.net">contact@datastewards.net</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b38c48e23e27" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[In Search of a Vision: The EU’s Approach to Tech Sovereignty Needs to be Steered by Clearer Public…]]></title>
            <link>https://medium.com/@TheDataTank/in-search-for-a-vision-the-eus-approach-to-tech-sovereignty-needs-to-be-steered-by-cearer-public-c88f85cc86f5?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/c88f85cc86f5</guid>
            <category><![CDATA[data-stewardship]]></category>
            <category><![CDATA[democracy]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[tech-sovereignty]]></category>
            <category><![CDATA[digital-sovereignty]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Thu, 11 Jun 2026 08:50:08 GMT</pubDate>
            <atom:updated>2026-10-12T12:30:18.155Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>In Search of a Vision: The EU’s Approach to Tech Sovereignty Needs to be Steered by Clearer Public Interest Outcomes Rather than a ‘Race’ to the Unknown</strong></h3><p><strong>Anna Colom</strong>, <strong>Marta Poblet</strong>, <strong>Paulina Behluli </strong>| The Data Tank</p><figure><img alt="An archival image of a medieval dinner scene. On one side, there is a Penelope (from Greek mythology) weaving, but instead of thread, it is a board of binary code. On the other side, there is a a drink spilling and other figures looking distressed/chaotic." src="https://cdn-images-1.medium.com/max/640/1*siWb92z4oVW0DBGUOtB04A.png" /><figcaption>Nadia Nadesan &amp; Digit / <a href="https://betterimagesofai.org">https://betterimagesofai.org</a> / <a href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</a></figcaption></figure><p>The European Commission finally published its proposed <a href="https://ec.europa.eu/commission/presscorner/detail/en/ip_26_1187">European Technological Sovereignty Package</a> last week to strengthen its autonomy and resilience as it develops its industrial strategy and digital ecosystem in a complex geopolitical landscape. In addition to the main document laying out the overall approach, the package includes the non-binding <a href="https://digital-strategy.ec.europa.eu/en/policies/open-source-strategy">EU Open Source Strategy</a>, the proposal for a <a href="https://digital-strategy.ec.europa.eu/en/policies/chips-act-2">Chips Act 2.0</a> to strengthen the semiconductor ecosystem and boost domestic demand, the also legislative proposal <a href="https://digital-strategy.ec.europa.eu/en/policies/cloud-and-ai-development-act">Cloud and AI Development Act (CADA)</a> to address cloud dependency, and a <a href="https://energy.ec.europa.eu/topics/eus-energy-system/digitalisation-energy-system_en#strategic-roadmap-for-digitalisation-and-ai-in-energy">Strategic Roadmap for Digitalisation and AI in Energy</a>.</p><p>In this article, The Data Tank revisits its recent <a href="https://www.datatank.org/news-insights/new-policy-brief-whats-next-for-europes-digital-and-data-strategies">recommendations on the way forward for Europe’s digital and data strategies</a> through the sovereignty lens proposed in this package. The focus on open source, the more explicit mention of environmentally sustainable measures to developing the European cloud computing capacity are welcome signs on the way forward. Yet, a clearer vision that is based on public interest tangible outcomes, as we propose below, is needed to steer Europe’s approach to digital sovereignty. Clarity on outcomes informs infrastructure and policy, and not the other way around. A vision based on public interest principles makes it more obvious why open digital ecosystems need to be accompanied by capacity and infrastructure that enable interoperability, decentralisation and a diverse supply chain. Clarity on outcomes also elucidates why a<a href="https://medium.com/@TheDataTank/beyond-digital-sovereignty-a-polycentric-digital-future-for-europe-0c2224bddd0c"> polycentric, decentralised approach to governance and democracy</a> is required for a thriving resilient innovation ecosystem with a public mandate or social license. This approach also means that agency and resources are devolved beyond the figure of the state, avoiding concentration and lock-in, to a wider range of layers vertically and horizontally, ensuring a dynamic and participatory approach to data governance and innovation. Finally, we examine the development of quantum technologies and infrastructure in Europe as a lens to illustrate our arguments.</p><h3><strong>Public-interest outcomes as the driving vision</strong></h3><p>The package by the European Commission proposing an approach to tech sovereignty is a step forward in many ways–such as in its emphasis on openness and environmental sustainability–, but the main document starts with the premise that Europe needs to reclaim its role in the ‘global race for geoeconomic power’. Joining a ‘race’ is a strategy without aim, destined to get lost without a compass. This premise adopts the language of the very same powers that Europe seeks independence from, namely Big Tech and its entanglement with US policy. Perhaps a more useful first step in defining an approach to sovereignty is to base it on a vision that is outcome-based and guided by public-interest aims. This is no denying that Europe has defined an overall vision through its strategic agenda and related policy development, from the European Green Deal, to the Democracy Shield or the Union of Skills. Yet, the aims of the Tech Sovereignty package need better alignment with these.</p><p>Clarity on outcomes is important because it can help design and prioritise policies so that investment and procurement are targeted towards those outcomes through approaches that will be sustainable and resilient. As The Data Tank has argued before, these should be public interest tangible outcomes that serve societies and the planet. For example: (i) aligning Europe’s computational strategy with binding targets to reduce absolute emissions in line with Europe’s 2030 climate goals as we have recommended in <a href="https://www.thegreenwebfoundation.org/news/tell-the-eu-keep-ai-within-planetary-boundaries/#acknowledgements">our response to the CADA consultation in 2025</a>, (ii) designing an industrial policy, cloud and AI computing strategy based on prioritising technologies that can improve science and health within principles of sufficiency, rather than the large scale required for consumer generative AI, (iii) designing policies, diverting funds and building ecosystems that support knowledge commons, information integrity and public interest media, including small and medium-size outlets of diverse languages as we conclude in our recent report on <a href="https://cdn.prod.website-files.com/63bd953ecb882b8e559c1614/6a0b38fae86b009d3027b07e_LitReview%20Media%20and%20GenAI%20DataTank%20May18.pdf">media, democracy and GenAI</a>; (iv) aiming for bottom-up innovation, decentralisation and diversification to avoid power concentration and lock-in. An outcome-based approach to tech sovereignty, then, would mean, ensuring autonomy and resilience and a digital and industrial strategy designed to improve and advance health and science; to build a thriving and diverse competitive economies based on localism, diversification and sufficiency principles rather than top-down concentration; to strengthen information integrity and the media ecosystem; to establish the foundations for resilient democracies, including the protection and fostering of civic agency, social justice and rights; and to ensure environmental protection and climate mitigation and preparedness. Starting with the outcomes, and the services they require, is a step that precedes building and regulating the technologies, infrastructure, data and ecosystems required and addressing vulnerabilities and dependencies that threaten these public interest outcomes.</p><h3><strong>Developing the capacity to leverage and govern data infrastructures</strong></h3><p>In addition to gain and maintain control over data infrastructures, Europe needs, across sectors, the skills and competences to build, run and govern data ecosystems. As The Data Tank has <a href="https://www.datatank.org/news-publications">argued in previous articles</a>, data stewardship as a strategic set of skills and competences and a function across organisations and sectors is essential to ensure a public interest-first approach to governing data and data ecosystems. The principles of data stewardship as per our proposed approach, ‘Act, Collaborate and Protect’, provide an important governance framework for advancing the EU’s technological sovereignty objectives. By promoting the responsible use and sharing of data, fostering cooperation among public and private stakeholders, and ensuring robust protections for individuals and communities, data stewardship supports the creation of trusted data commons that underpin Europe’s digital resilience, innovation, and competitiveness. As we mention in our latest <a href="https://www.datatank.org/news-insights/new-policy-brief-whats-next-for-europes-digital-and-data-strategies">policy brief</a>: ‘Europe’s emphasis on technological infrastructure, while important, must now be matched by an equivalent investment in human infrastructure. Data stewards, community intermediaries, and public-interest technologists are needed to bridge the gap between data supply and public demand’.</p><p>The Data Tank welcomes the emphasis on supporting SMEs, through mechanisms such as the European Competitiveness Fund (ECF) to channel and leverage investments. Clarity on the public interest outcomes of sovereignty will help decide the approach to supporting local ecosystems through funds and resources. Developing the data stewardship capacity to leverage and govern data infrastructures should be part of the mentioned ‘support to build advanced digital skills’.</p><h3><strong>Reiterating our call for polycentric democratic governance</strong></h3><p>The proposal by the EC hastens to clarify that its approach to sovereignty does not refer to isolation or protectionism. And while the use of the word ‘sovereignty’ as such to refer to strategic independence from tech oligarchies or unpredictable geopolitical dynamics might come and go, its use has deep connotations with state control, not necessarily aligned with the EU’s subsidiarity principles or democracy that respects sub-national and local constituencies. For this reason, as well as in alignment with a public interest outcomes-based approach to tech sovereignty, structurally embedding polycentric democratic governance is key not only from a point of view of values but for a coherent strategy and policy design.</p><p>The main document by the Commission refers to taking a ‘human-centric approach’ that upholds ‘EU values’ including ‘human oversight, respect for fundamental rights, inclusivity, equality, non-discrimination and accessibility, in line with the European Declaration on Digital Rights and Principles’. A human-centric approach is referred to as a lever to the goal of ‘[r]einforcing the ‘European way’ to technological sovereignty’ but we propose looking at it as an outcome too for clarity on the way forward.</p><h3><strong>Quantum innovation as a case in point</strong></h3><p>While not included in the EC’s package, tech sovereignty is already emerging in other ways on European soil through, for example, through the European Quantum Communication Infrastructure — <a href="https://qt.eu/ecosystem/quantum-communication-infrastructure">EuroQCI</a>. In this regard, quantum communication infrastructure is polycentric by design: quantum key distribution is point-to-point and physically grounded, and quantum communication networks create a mesh of links between specific places. For example, as reported in the <a href="https://journalismresearch.circle.so/c/next-internet-observatory/secure-internet-observatory-signal-5-may-26-7a146a92-ab46-4b23-9103-6f31f03a8caa">Secure Internet Observatory of the Media and Journalism Exchange network,</a> the <a href="https://www.cesga.es/en/portfolio-items/iberianqci-3/">IberianQCI</a> project will connect <a href="https://www.it.pt/Projects/Index/4993">Portuguese</a> and <a href="https://www.cesga.es/en/portfolio-items/iberianqci-3/">Galician</a> infrastructures in Valença do Minho and Vigo with terrestrial fibre backbone links, to be further connected with several peninsular cities through satellite links and optical ground stations (OGS). Through <a href="https://qt.eu/ecosystem/quantum-communication-infrastructure">EuroQCI</a> and Iberian QCI, networks will be increasingly federated, reflecting the distributed quantum competences and skills sitting in many European universities, research centres and small regional hubs.</p><p>This is the part of the stack whose native architecture already matches the tech sovereignty it has to realistically aspire to: distributed, federated, bottom up, close to local but interconnected nodes of knowledge and skills. Yet, it sits on a separate track. And there may be relevant strategic reasons for that. What is missing, in our opinion, is treating the distributed, granular layer as a key component of tech sovereignty.</p><p>This quantum track is also led by clearer outcomes as a key pillar of <a href="https://ec.europa.eu/commission/presscorner/detail/en/ip_20_2391">EU’s cybersecurity strategy</a>, including security, safety and resilience for EU, national and local infrastructure and governments, as well as solid connectivity for businesses and individuals. And despite the abstract nature of quantum physics as a complex science, it is led by tangible outcomes in more concrete and local ways, serving as a material map that mirrors the polycentric web of governance and the collaborative ecosystem approach that we propose.</p><p>Ultimately, our recommendation is that instead of framing tech sovereignty as a ‘race’, strategies and policies should be steered by clearer public interest outcomes, aligned with evidence, with a polycentric approach to governance and democracy, and with the wellbeing of people and the planet for designing effective and sustainable policies and infrastructure.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c88f85cc86f5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Media, Democracy and Generative AI: A Critical Juncture]]></title>
            <link>https://medium.com/@TheDataTank/media-democracy-and-generative-ai-a-critical-juncture-40aaa93b8cd4?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/40aaa93b8cd4</guid>
            <category><![CDATA[democracy]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[media]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Wed, 20 May 2026 04:23:30 GMT</pubDate>
            <atom:updated>2026-05-20T04:23:30.295Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Anna Colom,</strong> The Data Tank | <strong>Marta Poblet, </strong>The Data Tank</p><figure><img alt="Sam Altman watches over a family consuming his AI. The living room vignette is framed by power tool imagery. The words “Behaviour Power” are the top of the artwork, framing the intent of the piece." src="https://cdn-images-1.medium.com/max/1024/1*3VTqFUeiHl4VXglqSZ9xyQ.jpeg" /><figcaption>Bart Fish &amp; Power Tools of AI / <a href="https://betterimagesofai.org">https://betterimagesofai.org</a> / <a href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</a></figcaption></figure><p>The irruption of Generative Artificial Intelligence (GenAI) represents a watershed moment for public interest media and democracy. As we argued in a previous <a href="https://medium.com/@TheDataTank/strengthening-public-interest-media-in-the-age-of-genai-976ff27698e4">article</a>, the extractive and concentration-based practices of big Generative AI providers, coupled with inadequate regulation and a media ecosystem already weakened by existing trends towards platformisation, have implications for the sustainability and integrity of public interest media and knowledge ecosystems, key pillars of our democracies.</p><p>As part of the ‘Strengthening Media in the Age of Generative AI’ action-research and dialogue, The Data Tank conducted a <a href="https://cdn.prod.website-files.com/63bd953ecb882b8e559c1614/6a0b38fae86b009d3027b07e_LitReview%20Media%20and%20GenAI%20DataTank%20May18.pdf">literature review</a> which included a total of 221 academic, industry, and policy sources (you can read more details about the methodology in <a href="https://cdn.prod.website-files.com/63bd953ecb882b8e559c1614/6a0b38fae86b009d3027b07e_LitReview%20Media%20and%20GenAI%20DataTank%20May18.pdf">the report</a>). The aim was to map the challenges that Generative AI poses to public interest media in Europe –and to small and medium-sized organisations in particular– and the ways forward being proposed to address them.</p><p>The challenges and the potential solutions are multiple and interconnected. Based on our thematic analysis of the literature, we present them as diagnosis of the current context on the one hand, and as potential ways forward on the other, acknowledging, however, that there are no such neat or binary distinctions: the limitations of some proposed solutions can amplify the challenges identified. For example, the review finds that bilateral licensing deals between large media publishers and large AI corporations do not help small and independent media and do not solve the issues with misattribution and unreliability of Generative AI outputs (<a href="https://www.ebu.ch/Report/MIS-BBC/NI_AI_2025.pdf">BBC and EBU, 2025</a>).</p><h4><strong>Foster media sustainability and information integrity</strong></h4><p>As we write in the report, the challenges driven by Generative AI are multiple but, importantly, ‘<strong><em>are seen to affect the whole political economy of the media and information ecosystem, in a way that cannot be fixed by technical solutions, literacy and skills, or bilateral licensing deals alone</em></strong>’.</p><p>The impact of AI crawlers and content extraction without compensation is taking place at scale while audiences are also driven away from the media where their content is sourced to AI-summaries and multimodal tabs in AI assistants. This scraping is also placing a strain on the technical infrastructure of media outlets, which smaller organisations are less resourced to handle. This is impacting on the sustainability and the business models of media organisations.</p><p>At the same time, the quality of public interest content and the knowledge ecosystem is threatened as Generative AI assistants are inherently error-prone as they operate on probability-based inference rather than cognitive capabilities, unable to reliably provide the nuance, timing, context, attribution and verification that are essential to journalistic content and public interest media. The editorial gatekeeping processes and oversight mechanisms that exist for public interest media have its flaws, but provide layers of assurance and mechanisms for transparency and redress. By contrast, AI systems bypass the technical and governance standards and institutions that mediate the public sphere or, with perhaps less success, the increasingly complex or insular networked sphere. While this process of disintermediation had already worsened with social media networks and large platforms, AI systems are amplifying and complicating already concerning trends and their impact on democracy.</p><p>Evidence so far suggests that Generative AI is more likely to use English-language text, even in contexts where English is not the first language (<a href="https://linkinghub.elsevier.com/retrieve/pii/S0308596125000497">Brantner et al., 2025</a>), which, coupled with the sustainability issues, risks causing ‘news deserts’ (<a href="http://urn.kb.se/resolve?urn=urn:nbn:se:norden:org:diva-13697">Mansell et al., 2025</a>). Ultimately, as we write in the report: ‘<strong><em>The growing concentration of Generative AI power in a few hands, as well as concentration in ways information is generated (with English-centric and large national media favoured over more local or diverse sources), further erodes pluralism, equity, and diversity</em></strong>’.</p><h4><strong>Approaches to regain bargaining power</strong></h4><p>A central point in the literature and the diagnosis of the situation refers to what is considered to be a loss of leveraging power and an inadequate regulatory framework, not fit to deal with these structural impacts. On the loss of bargaining power, direct bilateral deals are out of reach for smaller media outlets, impact brand and trust and do not address issues of information integrity and pluralism. Yet, collective licensing such as the case of the <a href="https://dpcmo.dk/">Danish Press Publications Collective Management Organisation</a> exemplifies, media collaboratives or approaches to statutory licensing point towards potential ways forward to reclaim agency and negotiating power. On regulation, proposals range from copyright reform to competition regulation, and include backstop arbitration or the establishment of levy taxes applied to Generative AI corporations, the revenue of which would be redirected back to public interest media actors, as proposed by <a href="https://openfuture.eu/blog/copyright-ai-and-the-limits-of-voluntary-licensing/">Open Future (Keller, 2025</a>) or as recently argued by Mistral (<a href="https://www.ft.com/content/d63d6291-687f-4e05-8b23-4d545d78c64a#:~:text=Mistral%20CEO%3A%20AI%20companies%20should,holders%20and%20bring%20legal%20certainty">Mensch, 2026</a>).</p><p>Collective action by media coalitions and public and philanthropic investment are viewed as essential. And while media, data and AI literacy for audiences and media professionals alike are necessary, they will not be sufficient on its own.</p><figure><img alt="The cover of the literature review by The Data Tank on media, democracy and Generative AI. Yellow background, the graphic D of the logo of The Data Tank in a scale of yellows and dark oranges in the middel, and the date and authors at the bottom." src="https://cdn-images-1.medium.com/max/396/1*G3xk43hkUhX7QsBJuOd5qg.png" /><figcaption>A <a href="https://cdn.prod.website-files.com/63bd953ecb882b8e559c1614/6a0b38fae86b009d3027b07e_LitReview%20Media%20and%20GenAI%20DataTank%20May18.pdf">comprehensive literature review by The Data Tank</a> on media, democracy and Generative AI</figcaption></figure><h4>New visions and a holistic approach for the way forward</h4><p>Our review of the literature sheds light on the complex interplay between Generative AI, the media ecosystem, the integrity of information and, ultimately, democracy. In trying to capture this complexity, a few overarching themes stand out.</p><p>The current changes driven by Generative AI technologies both reflect and result in power asymmetries rewiring how information is produced, interpreted, distributed and used, bypassing democratic mechanisms for shared knowledge, contestation and accountability. Beyond technical fixes and media literacy, there is the need for solutions to strengthen collective bargaining power, supported by regulatory initiatives and investment in public digital infrastructure. These holistic solutions require collective action and collaborative, inclusive ecosystems with resources that include and support regional, local, niche outlets.</p><p>Adaptive, modular, and scalable business models and strategies are needed, including diversification and approaches that create spaces of belonging and trust to keep audiences engaged. Yet, not enough is known at this point about what strategies will work. Successful approaches to economic sustainability require time, funding and visibility.</p><p>Public interest journalism and media remain essential to preserve information integrity, accountability, and the diverse tapestry of voices and experiences within our democracies. Democratic governance of media, AI platforms and technology providers would better align the existing tensions between public interest and the reduced agency that comes with AI system architectures.</p><p>A collaborative approach and a multiplicity of actions, from the micro to the macro level, including institutional imagination, will be the essential compass in this critical juncture.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=40aaa93b8cd4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Data Stewardship Bootcamp Takes Off to Milan: Lessons Learnt and an Engaged Community]]></title>
            <link>https://medium.com/@TheDataTank/data-stewardship-bootcamp-takes-off-to-milan-lessons-learnt-and-an-engaged-community-fa65e14447e1?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/fa65e14447e1</guid>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Mon, 20 Apr 2026 08:28:04 GMT</pubDate>
            <atom:updated>2026-04-20T08:28:04.315Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Paulina Behluli</strong>, The Data Tank</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ws5A94YXsIpKU3h5OP2zdA.jpeg" /></figure><p>The Data Tank in collaboration with Fondazione Cariplo, and as part of the <a href="https://datastewards.net/">Data Stewards Academy</a>, completed, in the words of participants, a ‘unique’ one-month hybrid Data Stewardship Bootcamp in Milan. The bootcamp brought together and trained 23 participants from across civil society, social enterprises, and local and regional public and non-profit entities. Covering the entire <a href="https://datastewards.net/offerings">Data Stewards Canvas </a>via a step-by-step approach, the participants worked with several use-cases starting with identifying the demand for data and all the way to measuring impact for data-driven projects or services. The bootcamp combined lectures, including by affiliated <a href="https://datastewards.net/faculty">guest faculty</a>, with expert-led field visits to data driven institutions active in Milan. With three days in person in Milan and three online sessions, the bootcamp brought together local data and AI voices, but also managed to tap into global ones representing different contexts.</p><p><strong><em>What makes each bootcamp unique?</em></strong></p><p>Our data stewardship bootcamps are designed using an agile methodology that allows us to continuously contextualise both findings and use cases to the specific audience and geographic setting. Data governance regimes and stewardship capabilities are inherently complex and highly contingent on local conditions — ranging from data collection practices and interoperability standards to legal frameworks for data protection and sharing.</p><p>For that reason, each bootcamp is intentionally iterative. We adapt content in real time, selecting and refining case studies that resonate most with participants and reflect the realities of the broader ecosystem in which they operate.</p><p>In Milan, this meant engaging deeply with a set of contextually relevant cases, including <a href="https://transformtransport.org/research/urban-mobility-metrics/lab-roma050-sustainable-urban-mobility-in-the-metropolitan-city-of-roma-capitale/">Sustainable Urban Mobility Metrics in the Metropolitan City of Rome</a>, <a href="https://www.istat.it/en/data/">Istat data and accessibility</a>, <a href="https://cuebiq.com/">Cuebiq</a> cases –a location intelligence company that began its journey in Milan and is now advancing data collaborations through geospatial and mobility data.</p><p>The bootcamp brought together seven distinguished guest faculty from Italy and beyond, each contributing a distinct perspective on data stewardship in practice. Daniela Paolotti (ISI Foundation) set the stage by highlighting the value of data reuse for the public good. Ludovica Paseri (University of Turin) guided participants through European legal frameworks for data sharing. Gianluca Misuraca (Inspiring Futures Europe) mapped the broader AI and data policy landscape. Brennan Lake (Cuebiq) demonstrated how geospatial data can power cross-sector data collaborations. Alex Hutchison (Forr Data) emphasized the importance of ethical data practices and impact measurement. Finally, Gary Leeming and Emily Rempel (Liverpool City Region Civic Data Cooperative) illustrated how participatory data stewardship models can generate tangible public value.</p><p>Together, this combination of local and international expertise exposed participants to multiple entry points into data and AI governance — from legal and technical infrastructures to participatory and impact-driven approaches. The result is not just knowledge transfer, but the cultivation of a new generation of data stewards equipped to navigate — and shape — an increasingly complex and evolving data ecosystem.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YhBYeaoQBRNK6g5rKwVMhQ.jpeg" /></figure><p>In addition to the guest faculty, we also had the opportunity to step out of the classroom and engage in interactive, expert-led field visits.</p><p>During our first day in Milan, we paid a visit to Accurat where the team from <a href="https://www.systematica.net/">Systematica</a> and <a href="https://transformtransport.org/">Transform Transport</a> presented some use-cases on conducting research and executing projects to inform sustainable mobility practices in Italy and internationally. During our second day in Milan, we visited the <a href="https://www.labsimurb.polimi.it/about-2/">Laboratorio di Simulazione Urbana Fausto Curti</a> part of <a href="https://www.polimi.it/">POLIMI</a> and were introduced to a range of data collection exercises that they had implemented on city planning.</p><p>We also had a chance to see some data collection devices and got to wear the ‘architect’ shoes and design the neighbourhood with an interactive data tool that the Labsimurb team developed. Finally, we ended our three day bootcamp with a visit to Gallerie D’Italia that allowed us to walk through 19th and 20th century Italy via art.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aN5Zwe1mtAYQIaQzNW3_yw.jpeg" /></figure><p><strong>Incubating Data Collaboratives</strong></p><p>Another distinctive feature of the bootcamp is its built-in incubation process for data reuse and collaboration projects. Drawing on participants’ backgrounds and interests, we identified three thematic clusters in advance — Social Innovation and Equity, Climate, and Culture — within which participants could develop concrete ideas for data reuse.</p><p>Throughout the bootcamp, participants worked in these groups to shape and refine early-stage concepts. This included exploring how pilot data collaboratives might, for instance, increase the demographic diversity of visitors to cultural institutions in Milan, or help bridge the gap between insight and action in advancing climate-just policies.</p><p>Importantly, this work does not end with the bootcamp. It continues through a series of pitch sessions scheduled for June 2026, where each group will present a more developed concept in a rapid, focused format. These sessions are designed to generate constructive feedback from peers, facilitators, and guest faculty, while also testing the clarity, feasibility, and potential impact of each idea.</p><p>Beyond reinforcing the practical application of the bootcamp’s core concepts, this incubation approach creates a structured pathway from learning to implementation — positioning selected ideas for further development, partnership building, and potential funding.</p><p><strong>Exceeding expectations</strong></p><p>We received great feedback from participants, even exceeding expectations for some. The bootcamp came as a positive surprise for some because of its innovative and condensed format, which participants hadn’t necessarily experienced before. The following topics of the programme resonated the most with them: “Moving from Insight to Action: Decision Intelligence” and “Matching Demand and Supply: Governance”. As it emerged from the group discussions, moving from insight to action seems to be one of the challenges among the organisations, hence ideas brought forward on data storytelling, decision theatres, and more resonated with participants. The governance of data collaboration initiatives is another challenge that was raised and our framework based on the 4Ps of data governance (<a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5506604">Purpose, Principles, Processes, and Practices</a>) also received good traction.</p><p>What participants had to say about the experience:</p><blockquote>“This bootcamp was a highly valuable experience, strengthening my ability to structure, manage and ethically valorize large volumes of high-resolution meteorological data. It also fostered meaningful connections with organisations working on the social dimensions of climate adaptation, opening up concrete opportunities for future collaboration.”</blockquote><p>Cristina Lavecchia, Managing Director at <a href="https://www.fondazioneomd.it/">Fondazione OMD</a></p><blockquote>“A very in-depth and intensive course on data stewardship. Also recommended for those who work in this field to “systematize” their knowledge.”</blockquote><p>Rainer Maria Baratti, Project Manager at <a href="https://www.actionaid.it/">ActionAid Italia</a></p><blockquote>“This course is a must for anyone looking to lead in today’s datafied world. It defines the Data Steward not just as a technical role, but as a strategic bridge between those who have data and those who need answers. I walked away with the complete picture.”</blockquote><p>Riccardo Naidi, Innovation Consultant at <a href="https://cgm.coop/">CGM</a></p><blockquote>“It’s been a really useful bootcamp, hope to join it again in the future.”</blockquote><p>Simone Licata, Administrative Employee at the <a href="https://www.comune.milano.it/">Municipality of Milan</a></p><p><strong>Next Steps: Data Stewards Network</strong></p><p>As the bootcamp concludes, this cohort will be added to our growing pool of (9000+ learners and 180+ alumni). They will have the chance to follow our bi-weekly “Data Stewardship Trends to Watch” sessions and engage with other fellow alumni. This vibrant active community of data stewards is one of the main reasons we continue to engage in such meaningful work. While advances in data and AI provide a strong foundation for future innovation and scientific breakthroughs, it is the human infrastructure and data stewardship community that makes them possible.</p><p>There is a lot of work being done behind the scenes on bringing new data stewardship <a href="https://datastewards.net/offerings">modules and bootcamps to life</a>, so stay tuned and <a href="https://datastewards.net/contact">updated</a>. Also make sure to sign up to <a href="https://zcmp.eu/qWGm">The Data Tank’s newsletter</a> to be up to date with our plans and forthcoming programmes!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fa65e14447e1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Year in Review: Building Strategic Data Stewardship Capabilities Across the Globe]]></title>
            <link>https://medium.com/@TheDataTank/a-year-in-review-building-strategic-data-stewardship-capabilities-across-the-globe-321128fb653a?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/321128fb653a</guid>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Fri, 19 Dec 2025 12:06:19 GMT</pubDate>
            <atom:updated>2025-12-19T12:06:19.252Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Paulina Behluli</strong>, The Data Tank</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iVRdAaP4ChhGNfciH-HIIw.jpeg" /><figcaption>Data Stewardship Intensive Course in Berlin, 2025 C: Sebastian Pfütze</figcaption></figure><p>2025 marked another year of rapid acceleration in the data and AI landscape. Despite significant investment in AI and data infrastructure for the common good, data access gatekeeping and monetization efforts intensified. While initiatives supporting AI for the common good, such as the <a href="https://www.currentai.org/">Current AI</a> initiative launched at the <a href="https://www.elysee.fr/en/sommet-pour-l-action-sur-l-ia">AI Action Summit</a> in Paris continued to emerge, the exponential growth of data and AI reshaped geopolitical relations and raised concerns about the diminishing influence of civil society.</p><p>In response to these developments, Europe advanced its regulatory agenda through initiatives such as the <a href="https://digital-strategy.ec.europa.eu/en/policies/data-union">European Data Union Strategy</a> and the <a href="https://digital-strategy.ec.europa.eu/en/library/ai-continent-action-plan">AI Continent Action Plan</a>. 2025 also marked the year the EU Data Act became fully applicable, establishing new rules for fair data access, portability, and use across sectors.</p><p>Against this backdrop, The Data Tank, in collaboration with The Governance Lab, continued to work towards a critical and often missing strategic role: navigating responsible data governance and access as foundational human infrastructure for AI development. We worked closely with civil society, private sector, and public sector organizations across Europe and beyond to ensure that access to data for responsible AI development remains a priority.</p><p><strong>Building the Data Stewardship Foundational Skills</strong></p><p><em>End of Data Stewardship Bootcamp in Turin</em></p><p>We started 2025 by ending a course that we successfully held in Turin, Italy. In January 2025, we held our virtual pitch sessions with bootcamp participants who presented their organizational data reuse strategies aimed at accelerating the common good. These pitch sessions were a final opportunity as such for the participants to meet with each other and bounce off ideas between them.</p><p><em>Data Stewardship Course in Switzerland</em></p><p>The Data Tank then joined forces with the Governance Data Lab, Swiss Data Alliance, <a href="http://opendata.ch">Opendata.ch</a> with support from Hasler Stiftung and hosted the first Data Stewards Intensive Course in Switzerland. The course was organized as a one-week intensive and took place in between two cities, Zürich and Bern. It brought together 21 participants across federal, cantonal, private sector, and civil society exploring relevant topics on data stewardship, decision intelligence, data spaces and collaboratives, and many more. Our updated data stewards canvas was the main toolkit that guided the course. The course was expert led (welcoming 14 incredible guest lecturers) and included 4 field visits to interesting institutions, including a public transport operator company owned by and located in the city of Zürich (<a href="https://www.stadt-zuerich.ch/vbz/de.html">VBZ</a>), a prestigious local newspaper (<a href="https://www.nzz.ch/neueste-artikel">NZZ</a>), Federal Chancellery in Switzerland (<a href="https://www.bk.admin.ch/bk/de/home.html">Bundeskanzlei</a>). We wrapped our five days with a special visit to the Hackdays in Bern event and a visit at the Bundesrat.</p><p><em>Data Stewardship Course in Maryland</em></p><p>Our partners at <a href="https://thegovlab.org/">The Governance Lab</a> then embarked on a journey to Annapolis, Maryland where they trained chief data officers across the state’s government in collaboration with the Maryland Department of Information Technology. This bootcamp aimed at strengthening responsible data use and re-use across Maryland. Participants used the course to develop a data reuse strategy for their department or agency.</p><p><em>Data Stewardship Course in Berlin | Second Edition</em></p><p>As part of our skills and training efforts, we ended the year with a very special bootcamp that took place in Berlin organized in collaboration with Bertelsmann Stiftung. This course was the second edition of this partnership and gathered 19 participants across civil society and social sector organizations in Germany and beyond. The bootcamp was organized as part of the <a href="https://datenatlas-zivilgesellschaft.de/de/start">Data Atlas for Civil Society</a> initiative pioneered by our partners at the Bertelsmann Stiftung. It was made special by the community of alumni and partners that we have built, where a lot of participants from our first edition recommended the course to their colleagues who joined us. The course was organized as a one-month hybrid course taking place in-person and online. With 6 distinguished guest speakers (industry experts), the course tackled the use and re-use of data for the common good within the civil society landscape. For this course we paid a visit to <a href="https://www.vodafone-institut.de/en/">Vodafone Institute for Society and Communications</a>, <a href="https://citylab-berlin.org/de/start/">City Lab Berlin</a>, and <a href="https://tacticaltech.org/">Tactical Tech</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GClTV6ILunsivWetI-P6kw.jpeg" /><figcaption>Data Stewardship Intensive Course in Berlin, 2025 C: Sebastian Pfütze</figcaption></figure><p><strong>Strengthening the Data Stewardship Community</strong></p><p>As we expanded our data stewardship skills development program, we further strengthened our existing network of more than 180 global data stewards. We held four virtual “Trends to Watch” sessions that provided a space for our alumni to meet and go over emerging data stewardship trends of 2025 delivered by our Co-founder Stefaan Verhulst. These sessions were organized exclusively for our alumni network and then made public for the wider audience. Working with a global community of data stewards is always exciting, as we hear views on data and AI that sometimes overlap, but at other times are quite different, depending on the part of the world they represent. What remains constant is the view that strategic and responsible data stewardship is needed for the landscape to continue to develop. Beyond the events, we also shared opportunities (funding and job opportunities) on a weekly basis to our alumni group.</p><p><strong>Looking Ahead</strong></p><p>As we enter a new year we already have a pipeline of bootcamps lined up, data stewardship community meetup and new programs that we will be launching. All of this will be shared via our brand new website dedicated to our data stewardship work solely. Our data stewardship journey has just begun, so stay tuned for future developments.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=321128fb653a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is this the end of Business-to-Government (B2G) sharing?]]></title>
            <link>https://medium.com/@TheDataTank/is-this-the-end-of-business-to-government-b2g-sharing-43da96ab668e?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/43da96ab668e</guid>
            <category><![CDATA[data-act]]></category>
            <category><![CDATA[data-reuse]]></category>
            <category><![CDATA[data-stewardship]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Wed, 10 Dec 2025 10:39:15 GMT</pubDate>
            <atom:updated>2025-12-10T10:39:15.380Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Is this the end of Business-to-Government (B2G) sharing? How the European Commission’s Digital Omnibus Confines B2G Data Sharing to a ‘Last Resort’ Option</strong></h3><p><strong>Stefaan Verhulst</strong>, Co-Founder GovLab and The Data Tank</p><figure><img alt="An illustration of three „pixelated“ cupboards next to each other with open drawers, the right one is black" src="https://cdn-images-1.medium.com/max/640/1*VQTokELWO7BIbneZQIQD0A.png" /><figcaption><strong>‘Classification Cupboard’ by Anton Grabolle / </strong><a href="https://betterimagesofai.org"><strong>https://betterimagesofai.org</strong></a><strong> / </strong><a href="https://creativecommons.org/licenses/by/4.0/"><strong>https://creativecommons.org/licenses/by/4.0/</strong></a></figcaption></figure><p>The European Union’s pursuit to create a single data market has always been a balancing act between fostering public interest goals and safeguarding private enterprise. <a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj/eng">The Data Act (Regulation (EU) 2023/2854</a>), which became applicable on September 12 2025, codified this tension, particularly in its Business-to-Government (B2G) provisions under Chapter V.</p><p>Initially, these provisions required data holders to share data with public sector bodies in cases of “exceptional need”, which was divided into two tracks: urgent Public Emergencies and non-emergency Public Interest Tasks.</p><p>However, the European Commission’s <a href="https://digital-strategy.ec.europa.eu/en/library/digital-omnibus-regulation-proposal">Digital Omnibus Package</a>, published last month, has signaled a definitive pivot. The core message: B2G data sharing is now being refined and confined as a measure of last resort. This narrowing protects the private sector but simultaneously creates a critical challenge: without a designated data steward within both the public and private sector this restrictive, haphazard approach will fail to build the trusted, long-term data ecosystems necessary to address emergency and non-emergency, systemic societal challenges.</p><h3>The Data Act’s Original Intent</h3><p>The initial B2G framework in the <a href="https://digital-strategy.ec.europa.eu/en/factpages/data-act-explained">Data Act </a>(Chapter V, Articles 14–22) established the mandatory sharing mechanism, intended to prevent the absence of vital data from hindering public policy and crisis response. The concept of “exceptional need” provided the trigger, categorized as:</p><ol><li>Public Emergency (Article 15(1)(a)): A high-urgency scenario (e.g., natural disaster, health crisis) requiring immediate data access, and the only case where personal data (properly anonymized or pseudonymized, per Article 18(4)) could be requested.</li><li>Public Interest Task (Article 15(1)(b)): A non-urgent task provided for by law (e.g., official statistics, post-emergency recovery), limited to non-personal data.</li></ol><p>Crucially, even this original structure included a large barrier to routine government access, especially for non-emergency public interest tasks. The public body had to demonstrate it had <strong>“</strong>exhausted all other means at its disposal to obtain such data, including purchase of non-personal data on the market by offering market rates” (Article 15(1)(b)(ii)).</p><h3>The Digital Omnibus Package: A Confined Mandate</h3><p>The Digital Omnibus Package (the “Omnibus”), designed to streamline and clarify aspects of the EU’s digital laws, narrows this interpretation by targeting the core trigger of the B2G mechanism.</p><p>While the Data Act used the overarching term “exceptional need” to cover both urgency and public interest, proposals within the Omnibus aim to explicitly amend and focus the mandatory B2G access regime predominantly on <strong>“</strong>public emergencies” (as reported in <a href="https://www.twobirds.com/en/insights/2025/eu-digital-omnibus-package-major-changes-to-the-data-act-proposed">recent legislative analysis</a>).</p><p>What the Omnibus does to the B2G Mechanism:</p><ul><li>Narrowing: By deleting current Article 15 and replacing it with a new Article 15 (a), the mandatory sharing obligation is now being confined to just <em>public emergencies, (“relevant and proportionate to responding, mitigating, or supporting the recovery from public emergencies”)</em> . Thus the legislative intent shifts further away from allowing governments to use the Data Act as a tool for public interest data acquisition. This strengthens the protection of corporate data assets and trade secrets (safeguarded under Article 19.3).</li><li>Focus on Crisis: It confirms that the Data Act’s B2G mandate is mainly an “emergency button” — a tool for immediate, existential crisis response — not a mechanism for medium-to-long-term societal improvement.</li><li>Microenterprises and small businesses would be able to request compensation when mandated to provide data during emergency situations, recognising that compliance imposes disproportionate costs on smaller actors. Larger data holders, by contrast, would continue to be required to share emergency-related data without charge.</li></ul><p>For data holders, this narrowing may provide legal clarity and reduce costs (20 million Euros <a href="https://digital-strategy.ec.europa.eu/en/faqs/digital-package">according to the European Commission</a>) but also amplifies the compliance challenge: every mandatory request is now, by design, a high-stakes, time-sensitive event.</p><p>It is also worth noting that this shift diverges significantly from the recommendations of the <a href="https://digital-strategy.ec.europa.eu/en/news/experts-say-privately-held-data-available-european-union-should-be-used-better-and-more#:~:text=The%20rapporteur%20of%20the%20High,Europe%20and%20Digital%20Europe%20programmes.">EU’s High-Level Expert Group on B2G Data Sharing,</a> which emphasized building <em>enduring and proactive</em> frameworks for unlocking privately-held data for the public interest. Rather than an emergency-only regime, the HLEG proposed national B2G governance structures, recognised data stewards, regulatory sandboxes, and mechanisms to scale data sharing across sectors — all aimed at avoiding exactly the kind of fragmented, ad-hoc “last resort” scenario that the Omnibus now seems to institutionalize.</p><h3>Data Stewardship: The Missing Infrastructure</h3><p>To overcome the challenges of this narrowed B2G regime and, more importantly, to create the data ecosystem needed for wider societal benefit, organizations must implement robust in-house Data Stewardship structures and functions.</p><p>As argued in <a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5351072">our previous analysis</a>, a Data Steward is an organizational leader or team empowered to create public value by strategically and responsibly providing access to data for re-use. This function is the vital governance layer that transforms the Data Act’s coercive mandate into a foundation for proactive engagement.</p><h4>1. Mastering the Last Resort Compliance</h4><p>A Data Stewardship function is the only mechanism that can render the strict Chapter V compliance operational:</p><ul><li>Centralized Contact &amp; Speed: Serving as the single, dedicated interface with the public sector’s Data Coordinator (Article 37), a Steward can ensure requests are routed and responded to within the tight Article 18 timeframes.</li><li>Proactive Readiness: A Steward can institutionalize technical preparedness by maintaining data inventories and pre-certifying data pipelines for anonymisation/pseudonymisation, eliminating the need to “scramble” when an emergency hits.</li></ul><h4>2. Building the Long-Term Collaborative Ecosystem</h4><p>More significantly, a Steward’s function is to look beyond the immediate demands of the Data Act’s “emergency button”. Long-term societal challenges — such as sustainable infrastructure planning, public transport optimisation, or climate modeling — require consistent, voluntary data sharing. These initiatives cannot rely on the burdensome, “last resort” criteria of the new proposed Article 15(a).</p><p>By demonstrating legal and ethical excellence in handling mandatory B2G requests, a Data Steward builds the trust and transparency required to negotiate voluntary, pre-competitive, data-sharing partnerships that address non-emergency public interest tasks. This proactive governance, rather than reactive compliance, is the true engine for creating a dynamic and reliable data ecosystem that benefits society as a whole.</p><h3>Conclusion</h3><p>The evolution of the EU Data Act’s B2G provisions, amplified by legislative proposals like the Digital Omnibus, establishes mandatory sharing as a confined “last resort” mechanism, now primarily reserved for urgent public emergencies.</p><p>This narrowing ensures private sector protection but places immense pressure on data holders to respond flawlessly during a crisis. Without a dedicated, proactive Data Stewardship framework, companies will be perpetually locked in a reactive, haphazard compliance scramble that jeopardizes both their legal standing and the critical public mission. Data Stewardship is the bridge — it ensures effective execution of the emergency protocol while simultaneously creating the trust and institutional infrastructure needed for the longer-term data collaboratives that Europe urgently requires to tackle systemic societal challenges.</p><p>(Thanks to Begoña Gonzalez Otero and Anna Colom for review)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=43da96ab668e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[‘Relevance in turbulent times: what’s the role of civil society and philanthropy in Europe’s data…]]></title>
            <link>https://medium.com/@TheDataTank/relevance-in-turbulent-times-whats-the-role-of-civil-society-and-philanthropy-in-europe-s-data-fc5340839c12?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/fc5340839c12</guid>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Thu, 04 Dec 2025 08:26:45 GMT</pubDate>
            <atom:updated>2025-12-04T08:26:45.255Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>‘</strong>Relevance in turbulent times: what’s the role of civil society and philanthropy in Europe’s data and AI strategies?’</h3><p>Author:<strong> </strong>Paulina Behluli | The Data Tank</p><p><strong>Acknowledgments:</strong> <em>A special thanks to the following individuals and organizations for their contributions and review for the following blog: Anna Colom (The Data Tank), Clara Ruthardt (Bertelsmann Stiftung), Elisabeth Nöfer (Stiftung Mercator), Kai-Ti Wu (European Citizen Science Association), Matthias Spielkamp (Algorithmwatch).</em></p><p>We live in turbulent times. On one hand the data and AI landscape has exponentially developed. On the other, they have unfolded in ways that consolidate power and reshape geopolitical relations, raising concerns about the diminishing influence of civil society. Building on its European Data Strategy, the European Commission has pushed forward the Data Union Strategy and the AI Continent Action plan aiming to make industry data ready for competitiveness and innovation, especially in AI development. However, we argue that more work is needed for these strategies to ensure data and AI benefit people and the planet. To respond to these developments, and as part of a series of events held in Europe to launch our Policy Brief, The Data Tank (TDT) convened a panel in Berlin with speakers representing civil society and Berlin-based philanthropic organisations active in the fields of democracy, data, and AI.</p><p>Hosted by Bertelsmann Stiftung, the event welcomed around 50 participants from across the public, private, civil society, philanthropic and academic sectors. It kicked off with a presentation by Dr. Stefaan Verhulst (Co-founder of TDT and <a href="https://thegovlab.org/">The GovLab</a>) on TDT’s latest <a href="https://www.datatank.org/news-insights/new-policy-brief-whats-next-for-europes-digital-and-data-strategies">policy brief</a> and its five key recommendations, namely:</p><p><strong><em>Recommendation 1. Articulate open innovation as innovation that is purpose-driven and avoids monopolistic capture: </em></strong>We suggest defining openness as guided by a shared, purpose-driven vision to prevent inequity and ensure that governance, funding, and knowledge flows avoid monopolistic capture and empower innovators of all kinds to create public value. ‍</p><p><strong><em>Recommendation 2. Prioritise high-impact and tangible purposes: </em></strong>We highlight the need to focus data-related investments on tangible societal challenges, such as: preventive health, skills and education, information integrity and civic engagement, and sustainable urban and rural development, to name a few.</p><p><strong><em>Recommendation 3. Build the missing human infrastructure</em></strong>: We emphasize the need to upskill people and organisations to strategically steward data governance and incentivise data reuse for the common good.</p><p><strong><em>Recommendation 4. Update data governance: beyond consent towards a public mandate and new licensing forms: </em></strong>We argue here that current consent-based approaches should evolve in a way that capture meaningful democratic governance and participatory approaches so that there is a clear public mandate or social license for data reuse.‍</p><p><strong>R<em>ecommendation 5. Adopt polycentric and participatory governance‍: </em></strong>There is a need to enable a dynamic and decentralised governance model in the context of the Data Union strategy and AI Continent Action Plan so that the planned data labs and data flows across data spaces benefit diverse and local innovation and avoid top-down control.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mR46HkGKA6b70XcwhxiOEw.jpeg" /></figure><p><strong>Strengthening Democracy, Empowering Local Actors</strong></p><p>The need for local, people and need-driven alternatives to data practices was also highlighted by the panel. Clara Ruthardt, Project Manager for Digitalization and the Common Good at Bertelsmann Stiftung flagged the changing nature of democracy in the digital space and the importance of empowering local and regional actors. Clara underlined the need to balance curbing harmful data practices with strengthening collective alternatives — reimagining data governance, platforms, and participation so that technology serves real human and societal needs. One way Bertelsmann Stiftung is pursuing this is by building bridges and translating, convening, and connecting civil society, policymakers, and local and international partners to strengthen alternative data governance models and enable the sovereign use of data for people, communities, and democracy.</p><p><strong>Strong Civil Society and Fair Markets for Better Alternatives</strong></p><p>Openness and democracy was a recurring theme in the panel, also highlighted by Elisabeth Nöfer, Project Manager at the Centre for Digital Society at Stiftung Mercator. Elisabeth emphasized the role of civil society in ensuring that the digital transformation is in line with democratic principles and openness, as well as a fair, equitable and competitive data economy to allow better alternatives to emerge. For Stiftung Mercator, then, opening up markets, striving for European digital sovereignty and protecting democratic values should go hand-in-hand in order to develop data and AI solutions that push for the common good.</p><p><strong>Citizen Science for a Participatory Future</strong></p><p>Kai-Ti Wu, Strategic Tech Development and Fundraising Officer at the European Citizen Science Association (ECSA) raised important points about empowering a citizen science approach for EU data and AI policies. ECSA, a non-profit network of over 500 organizations and individuals across Europe, is dedicated to advancing citizen science in Europe and beyond. Connecting communities, data, and institutions lies at the core of its mission. Citizen science empowers people to collect and use data — such as monitoring air quality — to inform policy, generate local insights, and promote inclusive innovation. By aligning technical and ethical infrastructures, ECSA advocates for the integration of citizen generated data, demonstrating how participation can become infrastructure in itself. ECSA’s view emphasised the importance of a democratic, human-centred data ecosystem, where civil society and philanthropy play an active role in shaping Europe’s digital strategy through collaboration and participation.</p><p><strong>Finding the Leverage for Civil Society</strong></p><p>Ultimately, an important question that was raised and resonated with the speakers and the audience was: what is our leverage as civil society and philanthropy representatives in light of European data and AI developments?</p><p>Matthias Spielkamp, Founder and Executive Director of AlgorithmWatch acknowledged that we indeed live in turbulent times and there is a need for the voices of civil society, media, and communities to be amplified so that data and algorithmic systems benefit the many, not the few. Matthias also brought forward food for thought on different channels that can be used to push for effective implementation of data and AI laws and regulations, drawing, for example, on his active membership in the advisory board to the German Digital Services Coordinator elected by the German Bundestag.</p><p>Overall, the main points raised by both speakers and audience participants highlighted a common thread: i) ensuring that Europe’s industrial and data union strategy is steered in a more participatory way; ii) including voices that are often absent from the conversation to prevent extractive practices by major players; and iii) promoting a more human-centric approach.</p><p><em>This event was organised as part of The Data Tank’s 2025 events series. Stay tuned for more insights from our roundtable in Brussels, which gathered representatives of the European Commission, the European Committee of the Regions, Wikimedia, and Open Future among many other key players from the private, public and civil society sectors!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fc5340839c12" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Strengthening Public Interest Media in the Age of GenAI]]></title>
            <link>https://medium.com/@TheDataTank/strengthening-public-interest-media-in-the-age-of-genai-976ff27698e4?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/976ff27698e4</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[democracy]]></category>
            <category><![CDATA[digital-sovereignty]]></category>
            <category><![CDATA[independent-media]]></category>
            <category><![CDATA[genai]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Tue, 11 Nov 2025 12:00:36 GMT</pubDate>
            <atom:updated>2025-11-11T12:00:36.556Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Anna Colom,</strong> The Data Tank | <strong>Marta Poblet, </strong>The Data Tank | <strong>Stefaan Verhulst</strong>, Co-Founder GovLab and The Data Tank</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*WYLaM3NfuHjW5ETrcDi7YQ.png" /><figcaption><strong>‘Knowledge Sweatshop’ by Leo Lau &amp; Digit / </strong><a href="https://betterimagesofai.org"><strong>https://betterimagesofai.org</strong></a><strong> / </strong><a href="https://creativecommons.org/licenses/by/4.0/"><strong>https://creativecommons.org/licenses/by/4.0/</strong></a></figcaption></figure><p><strong><em>The developments in generative and General Purpose AI (GPAI) systems have raised concerns about extractive data practices, market concentration by a small number of big tech providers, and pervasive weaknesses in data governance. These have implications for how public interest media, a key cornerstone of our democracies, sustain their content production and protect its integrity. The Data Tank is </em></strong><a href="https://forms.gle/7BCfaoiZWgd3RE3V6"><strong><em>launching a new project</em></strong></a><strong><em> to convene key stakeholders in the media sector to identify strategies — such as data collaboratives, knowledge commons, innovative data licensing or advocacy coalitions — that can leverage the power of small and medium media organisations in Europe vis a vis large GPAI providers. The goal is to ensure that the media not only serve the public better but also derive fair value from their public interest work.</em></strong></p><p>Independent and public interest media have for long been considered a key public good for societies, a pillar of democracy and accountable governance. Well funded public media systems have been found to consistently <a href="https://journals.sagepub.com/doi/10.1177/19401612211060255">correlate with ‘healthy democracies’</a>.<em> </em>However, the struggle for public interest media as a resilient sustainable public good is a long-standing one.<em> </em>Decades-long market fragmentation, the digital transformations turning the public sphere into multimodal networked spheres, and more recently, the entering into the mainstream of generative AI (GenAI) systems, and General Purpose AI more broadly, have kept public interest and independent media in a constant state of survival and reinvention.</p><p>The current trends in AI development -particularly large GenAI models- risk creating a vicious cycle for journalism and public interest media. These models are trained on high-quality journalistic content but the methods for capturing the data have been associated with exploitative and extractive practices of web scraping, routinely disregarding property rights, licensing, <a href="https://www.ebu.ch/news/2025/10/ai-s-systemic-distortion-of-news-is-consistent-across-languages-and-territories-international-study-by-public-service-broadcaste">quality, (mis)representation and bias</a> as GenAI models often use these data without providing provenance, context or integrity. Once deployed, they often <a href="https://www.theguardian.com/technology/2025/jul/24/ai-summaries-causing-devastating-drop-in-online-news-audiences-study-finds">draw audiences away</a> from the very outlets that produced this content by providing AI-generated summaries and news-like outputs directly to users. As a result, media organisations lose readership, along with the advertising and subscription revenue that sustains independent reporting. With fewer resources, their ability to produce quality journalism declines, which in turn reduces the availability of trustworthy content. This cycle threatens not only the financial viability of public-interest media but also the integrity of the AI systems that depend on their work, and ultimately, the integrity of the available knowledge to the public.</p><p><strong>Economic viability, knowledge integrity and democracy: a multi-faceted issue</strong></p><p>This extractive and unreliable dynamic threatens information ecosystems while rendering high quality information providers, like public service media, unable to leverage their power as key custodians of the integrity of these ecosystems. The issue at hand is therefore multifaceted. At one level, it is a business model challenge, as generative AI undermines the economic foundations that sustain independent journalism by diverting audiences and revenue streams. At another level, it is an information and epistemic integrity challenge, since the erosion of high-quality reporting weakens the supply of verified knowledge that societies rely on to make sense of the world. It is also a challenge to cultural and language diversity, as AI systems tend to prioritise dominant languages and mainstream narratives, further marginalising local voices and minority perspectives. Taken together, these dynamics pose a profound challenge for democracies, where pluralistic media ecosystems and trustworthy information are essential to informed participation, accountability, and social cohesion.</p><p>The rapid pace of AI development heightens the urgency of addressing this issue and of the need to establish sustainable frameworks before market dynamics become irreversibly entrenched. Governments and institutions in Europe and globally recognise the challenges above. A <a href="https://informationdemocracy.org/the-high-level-panel-on-public-interest-media/#assessment">collective statement</a> by <em>The High Level Panel on Public Interest Media</em> recently called governments to ‘invest in new models to support and safeguard free and independent media’, adding that this ‘means that public support and investment in public interest media must be increased both nationally and internationally.’ This followed on the 29th October, with 29 countries adopting the<strong> </strong><a href="https://informationdemocracy.us18.list-manage.com/track/click?u=084f0c6d2665f72b95ca748af&amp;id=4c99b514de&amp;e=58f6de195d">Paris Declaration on Multilateral Action for Information Integrity and Independent Media</a> during the High-Level International Conference on Information Integrity and Independent Media. The <a href="https://gfmd.info/adoption-oecd-recommendation-on-information-integrity/">OECD Council had already adopted a Recommendation on Information Integrity on December 2024</a>. The European Commission is preparing the <a href="https://www.europarl.europa.eu/legislative-train/package-european-democracy-action-plan/file-european-democracy-shield">European Democracy Shield</a>, and opened a consultation in March 2025 as a response to the ‘intensification and wider spread’ of threats to democracy and the reliability of news and information. Last summer, in August 2025, the European Media Freedom Act (EMFA) <a href="https://digital-strategy.ec.europa.eu/en/news/europe-strengthens-media-freedom-independence-and-pluralism-new-european-media-freedom-act">entered into application</a> across the European Union, building on the <a href="https://digital-strategy.ec.europa.eu/en/policies/revision-avmsd">revised Audiovisual Media Services Directive</a>, aiming to enhance the integrity of the internal market, and thus protect media pluralism and independence in the Union.</p><p>Media and journalist networks, academia and non-profit organisations are attempting to understand and address these challenges. For example, the Observatory on Information and Democracy released a <a href="https://observatory.informationdemocracy.org/report/information-ecosystem-and-troubled-democracy/">‘A Global Synthesis of the State of Knowledge on News Media, AI and Data Governance’ </a>including a chapter on ‘<a href="https://observatory.informationdemocracy.org/wp-content/uploads/2024/12/rapport_forum_information_democracy_chapitre_2.pdf">News media, Information Integrity and the Public Sphere</a>’. In February 2025, the International Fund for Public Interest Media launched the <a href="https://ifpim.org/news/new-partnership-launched-to-ensure-linguistic-diversity-and-broad-journalistic-content-for-ai-models">Global Media Trust </a>in collaboration with Current AI to foster linguistic and journalistic content diversity by supporting news content creators particularly in the Global Majority. Last month, the Open Markets Institute and ARTICLE 19 convened a two-day <a href="https://www.openmarketsinstitute.org/publications/the-future-of-democracy-speech-thought-sovereignty-and-power-in-the-age-of-platforms-and-ai">conference to discuss democracy and power in the ‘Age of Platforms and AI</a>’ and the <a href="https://www.mfrr.eu/mfrr-summit-2025/">Media Freedom Rapid Response Summit</a> was also held, addressing questions such as how to ensure the economic viability of the media in today’s world.</p><p>Networks like the Global Forum for Media Development (GFMD) are asking for urgent help to support media organisations. Smaller media outlets are particularly affected: their content is crucial for well-informed thriving societies and for cultural and linguistic diversity. Yet, they have the least financial and influencing power to face the extraction of Big Tech and General Purpose AI providers. As stated by GMFD’s Executive Director <a href="https://mailchi.mp/gfmd.info/mediadev-insider-impact-may-25">Mira Milosevic during the World Press Freedom Day</a> 2025: <em>“Our members, especially those small media, community-based newsrooms and investigative journalism networks, are facing this multifaceted, increasing, and accelerated crisis — not only in funding, viability, and relevance but now, with AI, survival. And this is not just another challenge; this is a completely new dimension.”</em></p><p>A range of stakeholders are exploring ways to make media more resilient and sustainable in the age of GenAI. For example, academic institutions like the <a href="https://www.aim4dem.nl/projects/">AI, Media and Democracy Lab</a> at the University of Amsterdam are exploring ways to ’stimulate the innovation of AI applications that strengthen the democratic function of media’. A <a href="https://arxiv.org/abs/2501.17299">research </a>collaboration proposed creating a participatory open-source LLM’ and a ‘Newsroom Tooling Alliance’ where members could decide which of their journalistic content to pool in a shared database and hold the right to choose how it is used. There are also a range of technical innovations being tried to build infrastructure for responsible data reuse, such as <a href="https://diff.wikimedia.org/2025/04/01/how-crawlers-impact-the-operations-of-the-wikimedia-projects/">measures of human and bot traffic </a>to better control where resources are being directed, as explored by Wikimedia, approaches to attribution control, measures for <a href="https://www.commonsdb.org/blog/commonsdb-a-foundation-for-transparent-digital-rights-information/">greater legal certainty</a> over the reuse of digital content, <a href="https://www.theverge.com/news/775072/rsl-standard-licensing-ai-publishing-reddit-yahoo-medium">developing standards on existing protocols</a> like the robots.txt files, or <a href="https://www.impactlicensing.org/">innovative licensing</a> to balance the public good of open data with the need to attach social impact strings to extractive actors. There is also scope for applying the principles of data commons and data collaboratives to media organisations. In the European Union, for example, the <a href="https://tems-dataspace.eu/">European Media Data Space</a> aims at fostering cross-border collaboration and data-driven ecosystems to make the European media industry more competitive.</p><p><strong>Convening power and collective solutions</strong></p><p>In addition to the vital task of protecting information and democratic integrity, how can media providers, especially smaller outlets with limited resources and leveraging power, stand up to the extractive approach of current generative AI systems?</p><p>Bigger media outlets, like <a href="https://apnews.com/article/openai-news-corp-a49144d381796df5729c746f52fbef19">News Corp</a>, <a href="https://www.theguardian.com/gnm-press-office/2025/feb/14/guardian-media-group-announces-strategic-partnership-with-openai">The Guardian</a>, <a href="https://www.lemonde.fr/en/about-us/article/2024/03/13/le-monde-signs-artificial-intelligence-partnership-agreement-with-open-ai_6615418_115.html">Le Monde</a>, or <a href="https://www.washingtonpost.com/pr/2025/04/22/washington-post-partners-with-openai-search-content/">The Washington Post </a>have reached strategic partnerships with GPAI providers but there are problems with this approach. Firstly, general AI assistants <a href="https://www.bbc.co.uk/aboutthebbc/documents/bbc-research-into-ai-assistants.pdf">still provide inaccurate answers </a>due to the way large language models work, <a href="https://www.sciencedirect.com/science/article/pii/S0308596125000497">even when they draw on journalistic content</a> and cite sources. Secondly, smaller media outlets, those providing investigative journalism or important public interest local content in a range of languages, are less likely to have the power to reach these deals.</p><p>At The Data Tank we think there is power in convening small and medium size organisations alongside stakeholders from all sectors to learn and deliberate on what solutions can be explored, whether technical, regulatory or focused on enhancing negotiating and narrative power by strengthening coalitions. We want to foster a collective learning process on what tools can be leveraged by the media sector to improve its power in front of Big Tech and GenAI development, its sustainability and the integrity of its content. As stated earlier, this is, ultimately, crucial for the resilience of our democracies.</p><p>We are therefore starting a first phase of a project funded by Adessium Foundation and we are <a href="https://forms.gle/7BCfaoiZWgd3RE3V6">calling for expressions of interest </a>to collaborate. Our project aims to (i) help media organisations to mobilise power through convening and tapping into collective intelligence; (ii) strengthen the media community by providing knowledge on potential tools, solutions and pathways forward; (iii) help media organisations to agree on a roadmap through collective power and a focus on solutions.</p><p><strong>Our call to action</strong></p><p>If you are a media organisation or network, an academic or member of a research institution, a representative of a regulatory body, of the private sector, of a civil society organisation, or a person for whom the integrity media matters and are interested in being part of this multidisciplinary convening, please register your interest and <a href="https://forms.gle/7BCfaoiZWgd3RE3V6"><strong>share your views here</strong></a>. We look forward to hearing from you!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=976ff27698e4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[‘If only we had access to data for…’: Exploring Data Stewardship and collaboration for…]]></title>
            <link>https://medium.com/@TheDataTank/if-only-we-had-access-to-data-for-exploring-data-stewardship-and-collaboration-for-aa298d0b787b?source=rss-14056a7befdf------2</link>
            <guid isPermaLink="false">https://medium.com/p/aa298d0b787b</guid>
            <category><![CDATA[data-stewardship]]></category>
            <category><![CDATA[philanthropy]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[csv]]></category>
            <dc:creator><![CDATA[The Data Tank]]></dc:creator>
            <pubDate>Wed, 05 Nov 2025 14:41:02 GMT</pubDate>
            <atom:updated>2025-11-05T14:41:55.730Z</atom:updated>
            <content:encoded><![CDATA[<h3>‘If only we had access to data for…’: Exploring Data Stewardship and collaboration for philanthropic data innovation</h3><p><strong>Paulina Behluli</strong>, The Data Tank | <em>A pre-csv,conf workshop in Bologna</em></p><p>In recent years, Europe — the informally known global digital regulator, has pushed forward many digital and data strategies trying to address a rapidly changing environment. However, as stated in our <a href="https://www.datatank.org/news-insights/new-policy-brief-whats-next-for-europes-digital-and-data-strategies">policy brief</a>, the European Data Strategy has yet to achieve a dual imperative: protect people from extractive and opaque data practices, and unlock data in a way that systematically advances societal goals. At The Data Tank (TDT), we think that addressing this challenge requires a collaborative approach involving all stakeholders.</p><p>As part of a series of events in Europe, The Data Tank organised a workshop in Bologna in collaboration with <a href="https://www.compagniadisanpaolo.it/it/">Fondazione Compagnia Di San Paolo</a>. The workshop was organised in the context of the <a href="https://csvconf.com/">Csv,Conf,v9</a> conference, an international community event on open data.</p><p>The workshop in Bologna centered around the topic of data for philanthropies and gathered representatives across philanthropic organisations, civil society, and academia. We kicked off by introducing our <a href="https://www.datatank.org/about">2025–2028 strategy</a> and recent <a href="https://www.datatank.org/news-insights/new-policy-brief-whats-next-for-europes-digital-and-data-strategies">policy brief</a>, which makes five recommendations for Europe’s next steps for its data and digital strategy. One of our recommendations includes the importance of data stewardship, the human infrastructure needed to make these data reuse ecosystems work. Data stewards are strategic data leaders across sectors that enable access to data in a systematic, sustainable, and responsible way. At TDT, we have trained over 180 global data stewards and created a community of practitioners that goes beyond borders and data infrastructures. We highlighted three key data stewardship principles: collaboration, protection, and action. Data stewards work with others to unlock the value of data by reusing it across sectors and industries and make sure that the data is ethically managed and protected, all while acting in proactively identifying partners who can do so. There are several use-cases on how philanthropies around the world have unlocked the value of data. We shared, for example, The King Baudouin Foundation’s <a href="https://because.eu/en/analyse">Social Profit DATATrust</a>, and <a href="https://www.mcgovern.org/our-work/data-solutions/grant-guardian/">Patrick J. McGovern’s Grant Guardian</a>, to name a few. We also highlighted other important initiatives, such as the <a href="https://www.data4philanthropy.net/">Data4Philanthropy</a>. Such initiatives do not only provide food for thought on how data could be incorporated into the grant making cycle, but also bring together important actors around the ‘virtual table’.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9yiSJmlHigbQ9quekct4hQ.jpeg" /><figcaption>C: Hilel Dridi</figcaption></figure><p>Participants were also able to hear from Hilel Dridi, Data Program Manager at Compagnia di San Paolo, who took the stage and presented the <a href="https://www.compagniadisanpaolo.it/en/projects/data-strategy-2023-2030/">data strategy</a> of Fondazione Compagnia di San Paolo 2023–2030. Fondazione Compagnia di San Paolo is one of the very few European philanthropies that have published a data strategy.<em> </em>A key point made by Hilel resonates with a key recommendation in The Data Tank’s policy brief, which is the need to prioritise high impact and tangible purposes when reusing data for the public interest. Fondazione Compagnia has already started doing their part on focusing their data-related investments on urgent and tangible societal challenges, paving the way for other philanthropies to follow.</p><p>The workshop was an opportunity for participants to share the challenges faced in relation to reusing data for the common good, as well as potential ways forward.</p><p>The challenges discussed are not unheard of in the data space, showing they persist in a context of rapid societal and technological change. They include: balancing privacy and security with data access and sharing, overcoming cost barriers associated with governance and total costs of ownership, balancing innovation and regulation, building trust and data literacy, and inability to keep up with the speed of technological evolution. Participants also highlighted a lack of multi-level coordination and siloed operations related to the high costs of governance.</p><p>The participants then turned to potential solutions to these challenges in the philanthropic and civil society sectors. Data standards, asking good research questions, creating consortia (including philanthropies, civil society, and librarians to these consortia), and finding innovative ways of sharing synthetic data instead of raw data to not breach any privacy concerns were some of the recommended approaches. Beyond these solutions, participants also highlighted the need for more investment on internal data literacy to better understand the potential of data reuse. Yet, the high cost of infrastructure (human and technological) remained a barrier towards achieving such ambitions.</p><p>The insights from this workshop reflected TDT’s diagnosis of the current state of affairs in data reuse and the steps outlined in our strategy in many ways. Even though there are many initiatives geared towards building the supply-side of data and technology, there is a need for more investment to cover the needs for demand, governance, and human infrastructure.</p><p>This workshop was organised as part of TDT’s 2025 events series. Stay tuned for more insights from Berlin and Brussels!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aa298d0b787b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>