Top of page

Category: Artificial Intelligence

Dr. Duncan Eddy, Stanford Research Fellow

Artificial Intelligence Safety: An Interview with Stanford Research Fellow Duncan Eddy

Posted by: Leah Weinryb-Grohsgal

In this interview, Stanford Research Fellow Dr. Duncan Eddy reflects on his professional background, shifting his focus on artificial intelligence safety standards, and what he is currently working on. He offers insightful observations from his work in AI safety and advice for future professionals. Dr. Eddy shared his experiences with the Library’s Artificial Intelligence Community …

Posted in: Artificial Intelligence, Computational Research, Interviews

A screen grab depicting the user stories where C2PA standards could be useful. Text reads, "User Stories Organized by Content Lifecycle Model Steps 1. Content and Records Creation 1.1 How can C2PA document tools, including AI tools, used for descriptive metadata generation? 1.2 How can C2PA identify and potentially limit access to datasets for AI model training based on if the dataset content was human-created, AI created or a mix of both? 1.3 How can C2PA identify and potentially limit access to content for AI model training based on Copyright status? 1.4 How can C2PA be used to document the creation process, including digital tools, for new content to establish the baseline provenance data? 1.5 How can C2PA be leveraged to establish the provenance of open access publications governed by CC BY 4.o license variations? 1.6 How can C2PA be incorporated into audio digitization preservation processes to document provenance and continuity? 2. Maintenance and Use 2.1 How can C2PA be used to document stewardship of specific instances of content? 2.2 What role can C2PA play in documenting image accuracy of historical maps? 3. Accessioning and Processing 3.1 How can C2PA be leveraged in the acquisition process if incoming content already has C2PA? 3.2 What role can C2PA play for documenting images of artifacts for determining origination? 4. Preservation 4.1 How can C2PA enable provenance chains for any non-C2PA files? 4.2 How can C2PA relate or document derivative files with their authentic original or preservation/access copies of the same content item? 5. Access 5.1 How can C2PA be applied to “certify” the accuracy and completeness of copies of records generated for online access and in response to research requests? 5.2 How can C2PA be used to ensure the authenticity of redacted records? "

New Community of Practice for Exploring Content Provenance and Authenticity in the Age of AI

Posted by: Isabel Brador

Today’s post is from Abbey Potter and Isabel Brador of the Digital Strategy Directorate and Kate Murray of the Digital Collections Management & Services Division here at the Library of Congress. Since January 2025, a new Library of Congress working group has been exploring ways to bring responsible AI together with digital preservation through an …

Posted in: Artificial Intelligence