Exploring the Richness of Culture and Technology

Wikipedia turned 25 this year, and Wikimania came to Paris to mark it. From 21 to 25 July, more than 1,200 Wikimedians gathered in person with 2,000 online under the theme Liberté, Équité, Fiabilité (Freedom, Equity, Reliability) to ask how the free encyclopedia stays trustworthy in the age of AI. Internet Archive Europe (IAE) came with a working answer.

Archive Citation Sources: evidential support, at scale

On Wednesday morning, IAE Programme Manager Beatrice Murch and Head of AI Daniel Erasmus presented Archive Citation Sources & Claims, a beta tool built on a simple observation: of the roughly 100 million outbound links in Wikipedia, only a fraction point to academic literature. The Internet Archive’s Turn All References Blue project fixed millions of broken links. Archive Sources asks the next question: can we help editors plug the holes in their citations — making published research more reliable, one missing source at a time.

The numbers behind the tool are considerable. It indexes 43.9 million open-access academic articles from scholar.archive.org, broken into 237 million passages, from which it extracts around 1 billion candidate claims that editors could add to Wikipedia. Paste in any Wikipedia paragraph, and the system retrieves relevant academic passages in under two seconds, then spends about half a minute matching specific claims to specific evidence, with an explanation for each match.

The crucial design choice is what the system looks for. “We’re not really interested in topic overlap,” Erasmus told the room. “We’re interested in evidential support.” A standard search finds articles about the same subject. Archive Sources finds passages that confirm, contradict, or update what a Wikipedia paragraph actually says. Fine-tuning the retrieval pipeline for that task lifted accuracy from 0.68 to 0.92 on the team’s benchmarks. In one live example, the tool flagged a Wikipedia claim that two monoclonal antibodies had been approved for Alzheimer’s treatment, and pointed to a paper naming a third.

The workshop lived up to Erasmus’s promise to “put the work into workshop.” Attendees tested the tool live, and two of them pasted in French and Spanish Wikipedia articles unprompted. It handled both, a promising sign for a tool tested so far only on English. Just as important, the whole pipeline runs on infrastructure in Europe, operated by the team itself, with no reliance on external model providers.

The questions from the room went straight to what matters: how should candidate citations reach editors? Talk pages, a standalone tool, integration with existing workflows? Those decisions belong to the community, and IAE is listening. One attendee working on biodiversity heritage asked whether other collections could be added. The answer was an enthusiastic yes.

Mark Graham, the man of many sessions

Internet Archive founder Brewster Kahle could not make it to Paris, but the Archive was well represented by Mark Graham, Director of the Wayback Machine, who seemed to be everywhere at once and several members of the Wayback team. On Thursday, Mark joined the panel “Collateral Damage? Human Creativity and Interaction in the AI Crawling Era” alongside Creative Commons CEO Anna Tumadóttir, where he described how the Internet Archive is working with news organisations on alternatives to blocking the Wayback Machine, such as rate limiting and purpose-limited access. On Friday he returned for “Bridging Institutions, Wikidata, and Generative AI: From Data Silos to Shared Infrastructure,” on how memory institutions and AI systems can share infrastructure rather than compete for it.

Mark also brought something brand new to Paris at the Preconference: Wayback Machine Labs, an experimental playground he started building himself only weeks before the conference. The Labs already hold more than a dozen working prototypes built on the Wayback Machine, with contributions from the Wayback Machine team and the GDELT Project. Among them: Perspectives, a way to explore how media covers a searched topic; FiveThirtyEightIndex, an index of over 38,000 FiveThirtyEight articles, datasets, and podcasts preserved after Disney deleted the site, and What Changed, a tool that finds government webpages taken offline or quietly modified. His invitation to the community is an open one: come play!

Wiki Fringe: the open door

In-person tickets for Wikimania sold out well before July, and a new registration process made them harder to come by. The Wiki Fringe, an independent, curated side event running alongside the conference, gave everyone else a way in: a space for discussions, demos, and experiments where Wikimedians and the wider open knowledge community could exchange ideas without an official badge. Initiatives like this keep the movement porous, and that matters. Free knowledge has never been an invitation-only project.

One mission, many time zones

Beyond the sessions and the networking, Wikimania did something no video call can: it put Internet Archive people from across the globe, and friends from mission-aligned organisations across the free knowledge movement, in the same room at the same time. Colleagues who usually work an ocean apart compared notes over coffee, in corridors, and late into the Paris evenings. The group photo took nearly as much coordination as a conference session. Archivists, it turns out, are easier to convene around an idea than around a camera. Worth every minute, though, when looking at the beaming faces.

Last week, in an ancient forest an hour outside Berlin, more than 500 builders, researchers, archivists, and artists gathered to ask the same question in a hundred different ways: how do we build a web that cannot be switched off, sold, or shut down. Internet Archive Europe helped bring it there.

Ten Years On, First Time in Europe

For the first time in its ten-year history, DWeb Camp left the redwoods of Northern California. This year it landed at Alte Hölle, a former Stasi recreation site now home to a collectivist community, set on 100,000 square metres of forest and meadow in Brandenburg. The event is co-hosted this year by the Internet Archive, the Department of Decentralization, and IAE, which is why our co-founder Julien Masanès and our Programme Manager Beatrice Murch spent the week there rather than reading about it afterwards.

The move mattered. While the first Dweb Summit was held in 2016, DWeb Camp began in 2019 at an abandoned mushroom farm south of San Francisco, when Internet Archive founder Brewster Kahle brought together the creative spirit of Burning Man and the hacktivist energy of Chaos Communication Camp. A decade later, the same spirit crossed an ocean to gather closer to the policymakers, researchers, and institutions IAE works with every day.

None of this happens without someone tending it in between camps. Wendy Hanamura has been DWeb’s connective tissue since 2016, linking the builders, funders, and policymakers scattered across the movement and keeping the conversation about shared values, markets, and technical standards moving forward year round. Her pitch for the movement has always been straightforward: a decentralised web should let many different projects succeed side by side, rather than funnelling everyone toward a single dominant platform, which is closer to what the web was meant to be before it consolidated. It shows in how DWeb Camp runs. Attendees writing about the Berlin edition were quick to credit her and the wider DWeb team with pulling off the move to Europe.

Nine Tracks, One Direction

This year’s theme, Root Systems, was more than a nice metaphor. Mycelium networks don’t have a centre and don’t ask permission. They share resources peer to peer, and they survive when any single tree falls. That is the model DWeb Camp is building toward: protocols instead of platforms, infrastructure that can’t be pulled up by a single actor, however powerful.

Over five days, programming ran across nine content tracks: Anti-Authoritarian Stack, Cultivating Tech for Food Sovereignty, Decentralized Design, Decentralized Hardware and Local Community Networks, Open Social Web, Peer-to-Peer and Local-First, Public AI, Solidarity Tech, and Sustaining Infrastructure, spread across more than a dozen stages, tents, and workshop spaces. The full programme ran from morning yoga to midnight campfires, with sessions on mesh networking and cooperative governance sitting alongside a children’s puppet parade and a camp talent show.

Lock the Web Open

Brewster Kahle opened the camp with a talk that set the tone for the five days that followed, closing with a line that stuck with everyone who heard it: “Let’s build a decentralized web and lock the web open.” It is a simple instruction with a hard edge. Open access to knowledge is not a default state the web slides into. It has to be built, defended, and locked in place before someone else locks it shut.

One of the week’s most talked-about moments had nothing to do with slides. Attendee Andre Kudra described sitting down with Bruce Baumgart, who built the Internet Archive’s original “Petabyte Box,” for a long conversation about consciousness and whether machines could ever possess it. For Kudra, exchanges like that explained why the camp format works better than a normal conference: once people are sharing meals and tents for days on end, the small talk runs out fast and the real conversations start.

Those conversations quickly transformed into constructive brainstormings on Kahle’s own project OnionPress. The project pairs a familiar WordPress dashboard with a permanent Tor address and automatic backup through the Wayback Machine, and it picked up real interest at camp, with people asking how to get involved almost as often as they asked how it worked. It is still early and rough around the edges, but it works, and it says something about the mood at Alte Hölle that a tool built around exactly this instruction found such an eager audience. 

No Egos, Just Roots

Julien came back from Alte Hölle struck by something harder to put into a session title: the atmosphere. Wherever he went and whoever he spoke with, the same thing surfaced. People were friendly, generous with their time, and entirely focused on solving real problems for the public good. No egos, no turf wars, no cliques. Just a few hundred people looking in the same direction and working out how to get there together.

That showed up in the smaller sessions too. Bart Delrue, who teaches electronics and ICT at Odisee, ran an accessibility clinic that skipped the theory in favour of hands-on triage of real projects people had brought with them, and gave a talk on RSS as, in his words, the original simple and resilient root system. Elsewhere, the Platoniq Foundation led an immersive role-play on who gets to govern and fund democratic technology, and joined a Sustaining Infrastructure panel with the Free Software Foundation Europe’s Matthias Kirschner and Metagov’s Liz Barry on what keeps community-run infrastructure alive once the initial excitement fades. None of it was solemn. One evening session, built around a stash of glow-in-the-dark paint, flipcharts, and a Fediverse cape, filled the Social Web tent with campers making button badges and dancing to a Lily Allen track.

Coalitions around decentralisation and digital rights often fracture along exactly the lines DWeb Camp avoided: competing standards, competing funding streams, competing claims to who got there first. This one built something more durable than any single protocol: a working culture that treats collaboration as the default.

AI as Root, Not Threat

Sessions in the camp’s Public AI track treated artificial intelligence as a tool that could accelerate a more decentralised, public-interest web: language models built for communities with little existing infrastructure, public AI projects designed in the open, conservation and agricultural tools grounded in shared, non-proprietary data. A demonstration of the BetterEdge Sovereignty Anchor, a confidential-computing approach to running servers outside traditional data centres, drew praise from both Kahle and Baumgart. Built this way, AI spreads capability outward instead of pulling it into fewer hands, one more root in the system rather than another centralising force.

Why This Is IAE’s Work Too

Everything Root Systems stands for sits close to what Internet Archive Europe pushes for in Brussels and Amsterdam: memory institutions that can collect, preserve, and provide access on their own terms, rather than terms set by whichever platform happens to hold the data this decade. A decentralised web and a legally protected right to digital memory are two routes to the same destination. We therefore encourage you to sign the Our Future Memory Statement that aims at pushing decision makers in the right direction. 

As regards DWeb, it runs on volunteers and a community that keeps showing up year after year.  Moving forward, nodes should continue to spread and we are looking to start a DWeb node in Amsterdam, so please get in touch at office@internetarchive.eu. More generally, if Root Systems is a movement you want to help grow, please check out the DWeb Principles and use the DWeb contact form.

Photo credit: Anton Tal

More than 30 public AI researchers from around the world gathered at the Internet Archive Europe Amsterdam headquarters on June 9 to open the Data Care Festival organized by the Inclusive AI Lab, and to discuss the role of archives in supporting culture and society.

The discussion almost didn’t happen as planned. On 9 June, Internet Archive Europe (IAE) opened its Amsterdam space for the Fellows Soirée of the Data CARE Festival, a four-day gathering organised by the Inclusive AI Lab. The listed moderator, Kirthi Jayakumar, was not able to travel to the Netherlands because of a technical error in the passport. One of the listed panelists, Franklin Ozekhome, pop culture architect and founder of Pop Culture Varsity travelling from Nigeria, also encountered entry barriers and couldn’t make it that evening. Inclusive AI Lab founder Prof. Payal Arora, stepping in to moderate, named it plainly: a visceral reminder of who gets to speak, and what kind of passport shapes access to the very conversations about power and knowledge that the evening was there to have. 

The Data CARE Festival

The festival’s theme, “Reclaiming Techno-Optimism: Building for Context. Culture. Community,” reflects the core work of the Inclusive AI Lab, founded by Prof. Arora at Utrecht University. The lab incubates researchers, practitioners, and civic leaders from across the Global South and North, working on the concrete conditions under which AI can serve communities rather than extract from them. 

When Archives Speak Back

The evening panel, titled “When Archives Speak Back: Power, Data, and AI Storytelling,” brought together IAE Programme Manager Beatrice Murch; Dr. Jaswina Elahi, assistant professor at Utrecht University and principal investigator on heritage-building among postcolonial migrant communities in the Netherlands; Vincenzo Scagliarini, head of research at Logotel and editor of the collaborative economy project Weconomy (Italy); and Chux Daniels, who leads transformative innovation programmes across Africa. Rana Kuseyri, responsible AI researcher at the Inclusive AI Lab, had opened the evening by framing the festival’s core commitment: not optimism as a mood, but hope as a moral imperative.

The conversation that followed was grounded in lives, not abstractions. Elahi, whose research traces the heritage of postcolonial migrant communities in the Netherlands, pushed expanded the definition of what counts as culture. Growing up Surinamese-Hindustani in the Netherlands, she described a childhood shaped by Bollywood films, Surinamese Hindustani radio, and a recurring question: but where are you really from? Her doctoral work examined how digital platforms were allowing Hindustani communities in the Netherlands to construct and transmit cultural identity — research met, early on, with the assumption that young ethnic minority internet users must be at risk of radicalisation. The actual finding was that they were using the internet to feel connected to their communities, their home countries, and their culture.

That distinction matters for what archives do and don’t capture. Elahi was precise about it: heritage is not a building or a monument. It is a song that only exists in relation to another person. It is the way a grandmother cooks that her grandchildren can attempt to learn, but will always make it differently. The body carries heritage and passes it on. The recipes communities are writing down now, the YouTube searches for ingredients that are no longer available, the heritage books being compiled: these are not the heritage itself, but they are activating something that might otherwise be lost. Data alone is not heritage. But in the right hands, it can give communities a voice they were never offered elsewhere.

Scagliarini brought a different example: a team of five engineers from different countries, working for Cisco on a creative project that eventually made it to the Venice Biennale. The team included a designer who couldn’t code. From a conventional business perspective, that was a problem. What actually happened was that she and the engineers spent a month in conversation before the first GitHub push, and she came out of it having learned a new language, not to replace her own practice, but to build shared knowledge. Data, Scagliarini argued, is something living: it can always be broken apart, rearranged, and re-interrogated, even across centuries. The responsibility is to keep rewriting it, not to take any version of the record as fixed.

Daniels traced this across a different scale. The growing global presence of Afrobeats, with the nice touch of Nigerian music playing through Schiphol airport on his last arrival, is one signal of a cultural confidence that statistics about tech governance don’t yet reflect. Eighty-five percent of the world’s population lives in contexts where the biggest tech decisions get made without them. The UK Prime Minister meeting with Apple and Google to shape AI governance is not the same as the communities in Kenya, one of the world’s biggest social media user bases, having any say in how those systems work. The same asymmetry applies to knowledge-making more broadly: innovation in agriculture, finance, and mobility is being led in the Global South by people working without the infrastructure constraints that lock the Global North into old models. M-PESA exists because banks wouldn’t go to rural areas. The most interesting AI work may be happening in places the dominant platforms aren’t looking.

Beatrice connected this to the practical stakes of what IAE and the Internet Archive exist to do. Truth, she said, is fracturing. The archive’s job, namely establishing what was said and what happened at a given point in time, matters more when that fragmentation accelerates. Democracy’s Library, the Internet Archive’s project to gather government-funded public information and make it freely accessible, is one concrete response. Beyond that, the work comes down to choices made under constraint: you cannot archive everything. The guiding principle is not to let perfect be the enemy of done, and to be honest about the challenges, which are real: lawsuits, the rising cost of storage, legal frameworks that differ across EU member states, and journalistic organisations restricting Wayback Machine access. Thirty years in, the Internet Archive is still here, and still going.

Why IAE Supports the Inclusive AI Lab

IAE is a partner of the Data CARE Festival because the questions it poses are ones we share. Who controls the record? What gets preserved, and what gets lost? When AI trains on cultural heritage, whose heritage counts? These questions shape what libraries, archives, and memory institutions can do, and what communities can access and build on. The panel on 9 June was, among other things, a demonstration that these questions are not rhetorical. Two of the people who were supposed to be in the room didn’t make it, because of where they hold citizenship. That is the context in which memory institutions operate, and the context in which inclusive AI must be built. And this is what shapes what our understanding of the past will be in the future.