Hello everyone,
Please join the Electronic Records Section for an event titled "Community-Centered Frameworks for Content Authenticity and Provenance of AI-Impacted Records" to be held on Monday, June 8th at 2pm EST/1pm CST. Registration for this virtual event can be found here!
The speakers discussion will center around a recent white paper Content Authenticity and Provenance in the Age of Artificial Intelligence in the LAMs Community. A more detailed description is included below. We highly encourage you to review the white paper (though prior reading is not required) and bring your questions!
>>>
Generative AI has destabilized the evidentiary foundations that archivists have long relied upon to establish the authenticity and provenance of records, undermining the institutional trust that archivists are charged with sustaining. In this session, co-authors Joshua Sternfeld and Kate Murray discuss their recent report, https://blogs.loc.gov/thesignal/files/2026/04/Call-to-Action-CAP-for-LAMs.pdf. Click or tap if you trust this link.">Content Authenticity and Provenance in the Age of Artificial Intelligence in the LAMs Community, situating its findings for an audience of electronic records practitioners.
The report argues that no single standard, tool, or institution can resolve AI's challenge to content authenticity and provenance. Instead, as with other major disruptions to the archival community, safeguarding robust CAP practices will require creative strategies grounded in human judgment and field-wide collaboration. Sternfeld and Murray will introduce the report's four core pillars as a practical roadmap for navigating both the technical and human-centered dimensions of AI's impact on cultural memory institutions.
To illustrate these principles in action, the presenters will address organizational planning and access. At the planning level, they will introduce a tiered CAP maturity framework modeled after the NDSA Levels of Preservation, offering archivists a structured approach for organizational self-assessment and incremental progress. At the access level, they will demonstrate how coupling generative AI models with open-source frameworks such as Model Context Protocol can simultaneously enhance verifiable information retrieval while introducing new integrity risks. In both areas, the presenters will emphasize the importance of treating CAP not as a fixed value set but as a dynamic, evolving practice grounded in situational context.
------------------------------
Amanda Garfunkel
Digital Archivist
NewYork-Presbyterian/Weill Cornell Medical Center
------------------------------