We are not alone

Allies, archives and infrastructure in the AI age

Archives are having a moment.

Might they provide a future revenue stream for journalism in an AI age? Perhaps — there’s certainly a lot of talk about their potential. But why they’ve suddenly become all the rage may be more telling about the future and how the journalism industry should see our role in it. And who we should be — metaphorically — getting in bed with.

Researching an archive used to involve time and tedium: a basic search interface on top, mountains of un-optimized and undigitized material underneath, sometimes requiring in-person visits to rifle through actual, physical boxes. Archives are dense, dusty (sometimes literally), and often challenging to navigate.

But archives hold immense value; the challenge was always finding the exact information you needed within a vast repository of unstructured sources. It just wasn’t economically feasible to build the systems to sift through that data.

Now it is.

I’ve spent a lot of time thinking about archives this year, not just in terms of how we make newsroom archives more accessible (and monetizable) stores of information, but also more broadly, in the context of libraries and other institutions. In my other role uptown, I’m working with Columbia’s Incite Institute on an oral history of the global spread of investigative journalism over the last 50 years. The ultimate home of an oral history is an archive, and one of the core decisions in building a large oral history project is choosing how to make the material publicly accessible for however future generations may access information.

How will AI change the way archives approach organizing and preserving information? There aren’t any clear answers yet (that I’ve seen) about what future archives will look like, but there is excitement about making their contents more accessible, now that it’s finally affordable to build systems that make that possible.

Newsrooms have the same public access goals, though maybe a little more urgently, given the business we’re in and the crisis we face. At their heart, news organizations are — or should be — public-interest information providers and repositories. That puts us in company with institutions like archives, as well as libraries, museums, encyclopedias, weather services and more — anyone who has an eye to future generations looking at their collections.

We’ve long held ourselves apart, though, and for good reason. We work much more in the moment; we’ve historically spent as much effort on delivering the news as we have in creating it, and have always had a tenuous relationship with government funding (varying greatly across countries and political environments).

But that was then, and this is now.

We’re increasingly getting out of the delivery business, both by choice and necessity. News as the what-happened-yesterday commodity needs, thanks to AI, to turn more into news-as-the-latest-updates-paired-with-historical-context. Which positions us more as public-interest information providers, and aligns our interests with other public-interest institutions — perhaps more than ever before.

Journalists are not great joiners. But we’re up against fast-moving, seismic forces and historically rich corporations — as noted by A.G. Sulzberger’s call for collaboration within the industry in a recent WAN-IFRA speech.

Having a say in what the future of news looks like will likely require not just that collaboration across newsrooms, but also outside them, with other institutions that want to shape a future with informed communities at its center — which is, after all, the mission. Right?

It will also require a different way of thinking about our role in this ecosystem, beyond creating content and distributing it. It might mean getting more involved in building technology, or joining forces in new ways with government-funded institutions. But we need to work quickly; the new information environment is already taking shape.

There are already people bridging those divides, reaching out of their newsrooms, and sometimes leaving them altogether, to collaborate on public-interest information systems from different angles. Here are a few:

Libraries

Terry Parris Jr. is a longtime engagement journalist and, more recently, librarian, building library-based newsrooms that work with residents “to produce the news and information they actually need.” He built up the engagement reporting team at ProPublica, then directed engagement at The City (now The City Reporter) — where he created partnerships with local libraries — before getting a degree in library and information science while serving as the Public Square Editor for the New York Times’ Headway project.

To Parris, “libraries and journalism share the same foundations:” they help people make sense of the world, and are grounded by values of information, access, service and trust. Since 2024, he’s been building the Library Newsroom Project, which aims to create community newsrooms in the 16,000 public library branches across the U.S. He spent the last year as a Stanford JSK fellow exploring how to replicate that project, and is now working with the Urban Library Institute and News Futures on piloting participatory and community-centered models for local news and civic information.

Digital Communities

Trei Brundrett, co-founder and former Chief Operating Officer of Vox Media, has spent decades at the intersection of media and technology. While still involved on journalism boards, he’s now building a web app for local communities, called Roundabout. The project is housed in New_ Public, a nonprofit R&D lab building new digital public spaces that “connect people, embrace pluralism, and build community.”

New_Public is focused on trusted local digital spaces as alternatives to the current slate of social media platforms. The organization’s co-director, Eli Pariser, has likened these spaces to libraries (in a talk at the Vatican, no less). While not a journalist, Pariser’s work has long focused on how technology and media serve democracy — he helped lead MoveOn.org, co-founded Avaaz.org and Upworthy, and coined the term filter bubble more than a decade ago. The team he leads includes alumni of major news organizations and tech companies, alongside civic technologists, researchers and entrepreneurs.

Open-Source Software and Protocols

ProPublica’s senior director of technology Ben Werdmuller came to journalism from the open-source software world. Now, he’s taking a year off to do the same Stanford fellowship Parris just finished, to more deeply explore the ideas he finds most important at the intersection of journalism, technology, and democracy. He explained the choice and its timing on his blog (where he writes often about what the changing information environment means for journalism): “We’re in the fastest era of technological change in decades. Consequently, we’re also in the fastest era of journalistic change — and possibly the fastest era of democratic change.”

To him, this period of change is the chance to usher in “the community-first software era,” and the end of reliance on one-size-fits-all platforms. Werdmuller has long been involved in the open social web movement — technologists working to build decentralized alternatives to traditional social networks — where recent developments are enabling a flurry of innovation.

In particular, a new generation of open source web protocols allow users to share, discover, and build relationships using the software of their choice, without locking them into any one company’s services. That can be platforms that look like familiar social networks (think Bluesky) or more ephemeral, short-term networks, far easier to build now with LLMs. Werdmuller sees a lot of potential in short-term networks, and is excited to explore if the reporting questions newsrooms typically ask on callout forms embedded in stories could instead become dedicated community spaces.

“The result, I think, will be better information, but also more trust and loyalty built with the people we serve,” he told me.

Public Infrastructure

Ivan Sigal ran Global Voices, a multilingual media initiative that focused on amplifying voices from communities around the world, for over 15 years. He stepped down in 2024 to co-found Modal Foundation, which supports efforts to create alternative public-interest digital infrastructure to the dominant for-profit systems. The foundation is building some of the protocols and platforms that have Werdmuller excited, including AT Protocol, which grew out of Bluesky and has more than 40 million users. Its largest project is Eurosky, which is working to create sovereign (and open) digital infrastructure in Europe for the social web.

Modal also collaborates with the Public AI Network, a coalition to build AI as public infrastructure, made up of industry researchers, academics, policymakers, and civil society organizations. Their stance is that AI is becoming essential infrastructure, like “electricity, water, roads, libraries, or the Internet itself.” The network’s contributors include former WIRED editor-in-chief Gideon Lichfield, now a fellow at Harvard’s Allen Lab for Democracy Renovation.

Public AI focuses on tech, policy and investment efforts that center public access, public accountability and permanent public goods from AI infrastructure. As Lichfield describes in an essay exploring the concept and its conundrums, think public libraries, CERN or DARPA — but for AI. And yes, they already have a pilot with libraries.

If this sounds similar to the previous section, that’s because the two movements often overlap. But they differ in philosophy. Public AI is centralized — run in the public interest by a body accountable to the public (like a government). In contrast, decentralization is at the core of the open web: anyone can build on an open protocol, anyone can audit open source code. Sigal’s Modal Foundation is working to bring the values of the two together as “open social tech in the public interest.”

Standards and Metadata

We’ve talked about Sannuta Raghu here before. She came up with the concept of News Atoms — embedding metadata at the sentence level — a momentous undertaking to help secure provenance as information travels. She heads AI for the newsroom Scroll India, and a few months ago joined the International Press Telecommunications Council to explore whether news atoms can become a news metadata standard for the emerging agentic web.

Boring as that may sound, standards matter. And the IPTC is the organization for them in the news world. Founded in the 1960s to “safeguard the telecommunications interests of the world’s press,” it has set global metadata standards for different file formats used in news media through the many technological transformations since. Much of its current work is focused on media provenance as a defence (they’re based in London, so we’ll use their spelling) against misinformation.

The standards world is a dizzying set of coalitions, councils and collaborations: IPTC is carrying on the provenance work of Project Origin, an effort started by the BBC, Canadian Broadcasting Corporation, Microsoft and the New York Times in 2018, and building on the recently established C2PA Content Credentials for visual media developed by the Coalition for Content Provenance and Authenticity, an alliance between Adobe, Intel, Microsoft and others.

The upshot: these are the rooms where tech companies and (mostly large) journalism institutions are collaborating on systems for moving traceable news data digitally.

Yes, this is a list. Yes, you should follow all of these people. (We do!) But mostly it’s to show that we are not alone, and that there are many places and people journalists can find common cause with in this era of swift change.

True, the people here are pretty exceptional, and there aren’t that many of them. They are pursuing different paths, but all toward the goal of creating new institutions and infrastructure for a better information environment. But the leading edge can be lonely, which is all the more reason to bolster their ranks: to share our ideas, our values and our mission with potential collaborators, and to do so soon.

(And if you have names to add to this list, please let us know — especially if they’re doing something with archives, which is my pet interest.)

We all want the same thing: reliable, traceable public access to information.

Let’s go make some friends.