The CAIR 2027 workshop (Curators in the loop: human-centred AI for the digital curation of cultural heritage collections) will take place as a half-day, in-person event co-located with the 21st International Digital Curation Conference (IDCC27) on February 9, 2027.

Digital curation of cultural heritage collections is increasingly mediated by artificial intelligence: metadata is generated by models, texts transcribed by handwriting recognition, entities linked automatically, descriptions drafted by large language models. Each step reshapes the curator's role; often silently, and rarely by design.

The prevailing framing of "human-in-the-loop" collapses to the curator as reviewer of model output, understating the interpretive labour, disciplinary craft, and community accountability that curation practice carries. Responding to the IDCC27 theme, "FAIR DO's: Centring People in the Stewardship and Curation of Digital Objects," this workshop convenes work that pushes past oversight toward a richer account of curator-AI collaboration, as a design problem, an empirical research problem, and a governance problem.

We invite extended abstracts presenting empirical studies, design frameworks, deployed workflows, and critical reflections that treat human-centred AI not as a downstream deployment concern but as a first-order research question for digital curation. Contributions bridging AI research, human-computer interaction, digital humanities, and cultural heritage practice are especially welcome, as are contributions from early-career researchers, practitioners without a traditional research publication record, and community-based scholars.

Submission Guidelines

All contributions take the form of an extended abstract prepared according to the IDCC 2027 submission guidelines.

Extended abstracts must have up to 2500 words and describe the work being proposed for presentation and its relevance to the workshop topics. Submissions must use the provided template on the DCC website.

Submissions are in English and must be submitted as PDF via EasyChair through the following link: https://easychair.org/conferences/?conf=cair2027

Accepted abstracts will be scheduled for 15-minute presentations followed by 5 minutes for questions, grouped across two paper sessions.

Submissions will be reviewed single-blind by at least two members of the programme committee, drawn from across the AI, HCI, digital humanities, and cultural heritage communities.

Review criteria: relevance to the workshop’s central concerns; novelty and significance of the contribution; methodological rigour; clarity of presentation; and, for empirical work, evidence of engagement with curator practice rather than curator-as-abstraction.

List of Topics

We welcome contributions across four broad areas. The examples below are indicative; other work engaging seriously with the workshop’s central concerns will be considered.

• Design frameworks and interaction patterns for curator-facing AI systems: cataloguing tools, metadata enrichment, description generation, transcription (HTR/ATR), entity linking and disambiguation, retrieval augmentation, and interfaces that make model uncertainty and provenance legible.
• Empirical studies of curator experience with AI-augmented workflows: user studies, ethnographies of AI adoption in libraries, archives, and museums, longitudinal evaluations, cross-institutional comparisons, and studies of skills change and professional identity.
• Methodological and critical reflections on curator agency, expertise, and craft in AI-mediated pipelines: theorising the “loop” beyond oversight and review, evaluation frameworks that go beyond model accuracy, and situating human-centred AI within the histories of librarianship, archival theory, and museum studies.
• Community, participatory, and governance approaches to AI in cultural heritage: engagement with source communities, ethical governance frameworks, questions of consent and attribution in AI training data, and skills development and capacity building for AI-augmented curation.

Committees

Program Committee

• TBA

Organizing Committee

• Tomás Alves, DAViD Lab / BRU-Iscte, Iscte
• João Pedro Oliveira, DAViD Lab / IT, Iscte
• Alexandra Saraiva, DAViD Lab / DINÂMIA’CET-Iscte, Iscte
• Elsa Cardoso, DAViD Lab / ISTAR-Iscte, Iscte

Publication

A brief workshop report synthesising the day’s discussions and the outputs of the collaborative activity will be prepared by the organisers and submitted to an indexed journal.

Authors of accepted abstracts will additionally be invited to publish their work, at no cost to authors, as a chapter in the workshop’s book of abstracts, published online with an ISBN and with open access.

At least one author of each accepted abstract must register for the workshop day of IDCC 2027 and present in person.

Venue

The conference will be held in Iscte - University Institute of Lisbon, Av. das Forças Armadas, 1649-026 Lisboa, Portugal.

Contact

All questions about submissions should be emailed to tomas.alves@iscte-iul.pt