The IMLS-funded Responsible AI in Libraries and Archives project has published four new case studies, released in September 2026, that show how its Viewfinder toolkit works on real AI projects. Viewfinder is a participatory toolkit for ethical reflection on AI in libraries and archives, organized around different stakeholder perspectives. It was developed by librarians and technology ethicists at Montana State University, the University of Montana, James Madison University, and Iowa State University, based on research conducted from 2022 to 2025. The case studies cover a useful range of practice:

1. Robin Camille Davis on AI-generated device descriptions
2. Joshua Finnell on an organizational values inventory
3. Hannah McKelvey on evaluating AI features in licensed library resources
4. Erin Wolfe on newspaper collection metadata

Together they show what values-driven AI decision-making looks like in everyday library work, from vendor platforms to digital collections. Alongside the case studies, the project offers an interactive web version of Viewfinder, a print-at-home version on OSF, and a demo video of a sample session recorded in August 2026. Explore the case studies and toolkit.

Archive Stories is an open-access website about working with creative and non-traditional archives. It is curated by Sara Salem and Mai Taha, designed by Frederick Kannemeyer, and funded by the London School of Economics and Political Science. Rather than offering a new definition of the archive, the curators focus on archiving as a creative practice and as a response to the limits of institutional archives, including national archives that have been destroyed or made inaccessible and colonial archives that reproduce colonial power. The site now gathers dozens of contributions, several of them in Arabic, and many of them trace histories of anticolonial and revolutionary solidarity. Topics include the Palestine Poster Project Archives, traces of PFLP cinema in Beirut, a community-built archive of Syrian cassettes, Chilean revolutionary arpilleras, and oral history traditions in Palestine. The site is designed to be explored without a set path, and the curators welcome new submissions.

Old Toronto turns more than 37,000 archival photographs from the City of Toronto Archives and Toronto Public Library into walking tours of the city. The tool builds on OldTO, the geocoded photo map originally created by Sidewalk Labs and later revived by Back Lane Studios. While OldTO showed what was photographed in a given place, Old Toronto actually generates walking routes of the city. Users pick one of 25 starting points and one of 12 historical eras, from 1788 to 2018. The Planner then generates an optimized walking route anchored by landmarks, with clusters of archival photos along the way. Because each era’s photographs are distributed differently across the city, each era produces a different route through the same collection. The project explores approaches to visualizing archival collections that bridge the organizational logic of the archive, arranged by provenance, date, and description, and the lived experience of a city, organized by place and movement.

The Prison Data Explorer is a project of Trans Pride Initiative (TPI), a Texas nonprofit, that presents live data from TPI’s correspondence with trans and queer people in prison, primarily in Texas, but also in other state and federal prison systems. The Explorer charts more than 28,000 letters exchanged with over 2,000 people, including the volume of letters sent and received each quarter and correspondents’ race, age, gender, and sexual orientation. It also documents more than 18,000 incidents of violence and mistreatment reported to TPI, each classified using TPI’s published definitions. Users can filter charts and search individual records. Companion pages track prison systems’ responses to TPI’s complaints and data for Prison Rape Elimination Act (PREA) auditors. The project offers a community-generated approach to documenting conditions in carceral facilities.

Two organizations, Local Contexts and Creative Commons, have collaborated to create guidelines that “share commitments to equitable knowledge sharing, proper attribution, and respect for diverse cultural and community protocols,” in support of cultural heritage institutions that are creating access to data for Indigenous collections. These guidelines go beyond the usual protocols of copyright, priming museum, library, and archive workers to consider ethical and legal issues of open access to Indigenous cultural heritage. The new guidelines

“advance practical understanding among cultural heritage practitioners of how open heritage interplays with the protection of Indigenous rights over their heritage and data. Following a hands-on approach, the guidance unpacks the key issues underlying equitable access to and reuse of Indigenous cultural heritage and maps out the practical steps institutions can take when engaging with Indigenous communities… [and] clarifies how the CC licenses and tools and Local Contexts Labels can work together towards greater cultural equity, supporting communities in sharing on terms that reflect their cultural authority, and supporting respectful cultural exchanges and intercultural dialogue. In other words, it aligns open licensing tools with Indigenous data sovereignty frameworks.”

The guidelines themselves and a webinar explaining an overview are available on the Creative Commons webpage.

In this blog post, Stuart Dunn explains the implications of a vote of the United Nations General Assembly to adopt the Equal Earth projection over the more commonly used Mercator map, which privileges the degrees of travel between to points on a sphere to surface area. In addition to explaining the differences between the two maps, Dunn also shares a link to a resource that visualizes how the differences have markedly altered the concept of relative size of different countries.  The author is also sure to point out that both projections, while necessary for different purposes, are not perfect:

Equal Earth certainly goes some way to addressing the historic visual imbalance [the Mercator map] created. However, and I will no doubt not be the only person to point out, it is not a “correct” or “accurate” visualisation, or even a “complete” one – no projection is. It is a set of mathematical abstractions and compromises which renders visible what is otherwise too huge to be visible. In that process, some combination of data – be it angles, surface area or symbology – will always be compromised. While the UN vote re-surfaces this debate and brings the matter back to public attention, Equal Earth is certainly not a panacea — a fact fully recognised by the authors of the paper in which it was originally proposed.

This post is a useful resource to explain to student the history of cartography as context for working with maps in the modern day, and it encourages those who read it to think critically about visualizing geographical data.

Alyssa Virker, with the Data Sitters Club, shares a very practical, no-frills post about topic modeling as part of its TL;DR series, which is aimed at people who “don’t know if digital humanities is right for you.” This post is ideal for introducing students to the method as a form of inquiry and exploration, in under 1,500 words. Rather than giving a step-by-step for conducting a topic model, this post explains what this particular mode of inquiry does and does not do, and that “topic modeling is a mere mortal swimming in ambiguity with the rest of us.”

The post includes links to to more dense theoretical frameworks and step-by-step guidance.

An online conference on November 16, 2026, hosted by Canadian HSS Commons, will discuss how digital humanities projects that “span multiple generations of scholars, infrastructures, and institutions” can teach future scholars about project management, transitions, and tools. The format of the conference will be pre-recorded videos, accompanied by live conversation.

Registration is free, but required by November 10, 2026. Registrants must be members of the Canadian HSS Commons group to participate.

SWIB26, the 18th Semantic Web in Libraries conference, will take place online November 16–18, 2026. Organized by ZBW – Leibniz Information Centre for Economics and the North Rhine-Westphalian Library Service Centre (hbz), SWIB brings together practitioners working on linked open data in libraries, along with participants from scholarly communications, museums, archives, and related fields. This year’s keynote is Mark Dingemanse on the interactional foundations of critical AI literacies. The program also includes 13 presentations, five hands-on workshops, and a community session on SWIB’s AI policy. The conference is free, and the livestream is open to all. Workshops run on Monday, November 16, and require a free SWIB forum account to register. Sessions are scheduled on Central European Time. Details are on the SWIB26 program announcement and the registration page.

The Global Digital Humanities Symposium invites proposals for its 12th annual event, taking place both virtually and in person May 2027. The Call for Proposals is open until November 18th, 2026. The organizers especially welcome submissions on the following topics:

• Multilingualism and language justice in DH

• Indigenous led DH work and community engaged collaborations in DH

• Globally situated human-AI collaboration and critique in DH

• DH approaches to the study of the environment and considerations of environmental justice in DH infrastructure and practice

• Relational, distributed, and/or creative frameworks and practices in global digital pedagogy

• Public and community-engaged DH

• Digital historical research and archives

• Minimal, material, and sustainable approaches to DH

• Surveillance, censorship, and/or data privacy in global contexts

• Open data, open access, and data preservation as resistance

• Borders, migration, and diasporas with an emphasis on the effects of warfare and conflict

• Feminist and queer perspectives in DH

• Disability justice and accessibility

• DH methods in interdisciplinary and cross-regional research

• Digital storytelling and artistic representations in digital spaces

• Computational literary studies

The schedule of events will be as follows:

• 10 May (Monday) – In-Person Symposium at Michigan State University (in English)
• 12-13 May (Wednesday-Thursday) – Virtual Symposium (in English, Spanish, and French)
• 15 May (Saturday) – In-Person Symposium at the University of Guelph (in English and French)
• Mid-May, Post Symposium – DH@Guelph Summer Workshops

DH Benelux 2027 will be held July 7–9, 2027, at the Université libre de Bruxelles in Brussels, with pre-conference workshops on July 6. The theme, “Where Is Digital Humanities?”, asks where DH does its work after LLMs: with whom, on what terms, and at what cost. The call is organized around three axes: DH’s disciplinary identity, its credibility with publics such as citizen scientists and heritage volunteers, and its place in a research lifecycle shaped by generative AI, including that AI’s environmental costs. Several suggested topics speak directly to library and archives work. These include cultural heritage institutions as research partners, digital archives and long-term preservation, reproducibility and citation when models are versioned or deprecated, and crowdsourcing and participatory annotation. The conference accepts proposals in English for oral presentations, lightning talks, posters, demos, panels, and workshops. All submissions are anonymized and go through double-blind review. Presentations must be in person, since no online option is planned. Proposals are due December 7, 2026. Full details are in the call for papers.