2026 AI4LAM Fantastic Futures: Trust in the Loop

Save the Date!

Preliminary Schedule, subject to change

• February 9, 2026: Theme, call for proposals and timeframe announcement

• March: Full submission instructions and submission portal opens

• April 6, 2026: Proposals Due 

• June 1, 2026: Decisions & notifications

• July 6, 2026: Conference Schedule released

• September 14, 2026: AI4LAM DC Chapter-hosted events, affiliate meetings, tours

• September 15, 2026: AI4LAM Member Forum and FF26 Workshops

• September 16-17, 2026: 2026 AI4LAM Fantastic Futures: Trust in the Loop

• September 18, 2026: AI4LAM DC Chapter-hosted events, affiliate meetings, Working Group meetings, tours

Trust in the Loop

Trust is a cornerstone of libraries, archives, and museums. Trust in the Loop extends the idea of “human in the loop” to focus on the work LAMs must do to ensure that AI-enabled services are built on trust, authenticity, and accountability.

The Program Committee welcomes proposals that explore how trust is challenged, maintained, or strengthened as LAMs experiment with, implement, and are impacted by AI technologies.

CALL FOR PROPOSALS 

Please prepare submissions to the Fantastic Futures 2026 program that respond to the topics and questions listed below. 

Topic 1: Implementations

We invite the AI4LAM community to share how AI tools and systems are being implemented in libraries, archives, and museums.

LAMs work within real constraints, including limited resources, legacy systems, complex data, and responsibilities to staff, users, and communities. Proposals should explore how organizations are making progress with AI while supporting ethical and responsible practices.

Strong proposals would address some of these relevant questions:

• What principles, frameworks, or governance structures guide AI investment and decision-making?

• What goals are driving AI implementation?

• How are risks identified and mitigated?

• How is success or impact measured?

• What challenges arise when scaling pilots into operational systems?

• What long-term benefits and costs are anticipated?

• What AI implementations are being pursued, reconsidered, or deliberately avoided?

Topic 2: Data

AI systems are data-driven, and responsible AI implementation depends on the quality, structure, and context of data.

LAMs steward vast amounts of unique and culturally significant data. However, this data is often messy, incomplete, multilingual, historically situated, or poorly represented in foundation models.

Proposals may address:

• Risks and benefits of using LAM data in AI systems

• Data preparation, documentation, and standards for AI use

• Reviewing and validating the quality of AI outputs

• Effective data structures and preparation methods

• Use of controlled vocabularies in AI workflows

• Shared performance benchmark datasets for LAM content and tasks

• Implementation of MCP or other approaches to manage external use of data

• Strategies to manage data scraping or approaches to navigate relationships with commercial AI companies

• Emerging policies, procedures, and governance practices for AI data use

Topic 3: People

AI adoption affects users, visitors, patrons, staff, and communities.

With widespread access to AI tools, audiences expect new services, and staff are exploring how AI fits into their work. Proposals should focus on how organizations support people both inside and outside the AI loop.

Questions to consider:

• How are AI tools used to engage users in new ways?

• What training or capacity-building programs exist for staff or users?

• How is staff expertise integrated into AI workflows?

• How are AI outputs reviewed for quality and appropriateness?

• How do staff experience and respond to evaluating AI outputs?

• How are communities and stakeholders engaged when LAM content becomes AI training data?

• How are expectations managed with vendors, leadership, boards, or oversight bodies?

Topic 4: Models & Infrastructure

LAMs have a long tradition of collaboration around professional practices, open tools, and shared infrastructure. AI presents new opportunities—and challenges—for collective work.

Relevant topics include:

• Shared infrastructure needs for AI in LAMs

• Long-term data storage, security, and transfer requirements

• Approaches for sharing data, models, and documentation across institutions

• Evaluating AI tools and models for LAM use

• Reporting incidents or issues related to AI operations in LAMs

• Estimating and managing long-term costs and resource use

Submission Instructions: 

Submissions for the 2026 AI4LAM Fantastic Futures: Trust in the Loop program are due Monday, April 6, 2026. A link to the submission portal will be shared closer to the due date along with more detailed instructions. The information below is for planning purposes and may change or expand. The Program Committee will accept the following types of proposals: 

• Lightning talk

• Short paper

• Long paper

• Panel

• Workshop

Selection criteria: 

Proposals will be reviewed by experts from the AI4LAM community. Final programming and format decisions will be made by the Program Committee.

For context, 150 proposals were submitted to the 2025 conference, and 80 were accepted. Proposals are welcome from both AI4LAM members and non-members.

Proposals will be evaluated based on:

• Relevance: Alignment with the conference theme and AI4LAM priorities

• Applicability: Usefulness to the AI4LAM community

• Clarity: Accessibility to a broad, cross-sectoral, international audience

• Quality: Strength of evidence, research, or practical experience

Additional format-specific considerations:

• Workshops should emphasize interaction and practical application

• Panels should represent diverse perspectives and expertise

Coming soon

• Full submission instructions and portal

• AI4LAM Futures Challenge Announcement and Innovation Awards

Sign up to the AI4LAM email list for updates.

The AI4LAM Fantastic Futures 2026: Trust in the Loop conference is co-hosted by the Library of Congress, the National Gallery of Art, and the Smithsonian Institution, with support from staff at these organizations and the AI4LAM DC Chapter.

The Program Committee is led by the co-hosts, with guidance and input from AI4LAM members, the AI4LAM Board of Directors, and experts from the libraries, archives, and museums (LAM) community.