What happens when archivists, librarians, and digital collections leaders from a wide variety of institutions—large research universities, small liberal arts colleges, public policy institutes—gain early access to a new AI-powered tool purpose-built to accelerate collection processing and transform access?

You get enthusiasm, ideas, and a glimpse of what’s ahead, as more diverse primary source collections become discoverable for the first time.

This summer, charter participants of JSTOR Digital Stewardship Services gathered to share early experiences with JSTOR Seeklight, our AI-powered tool for accelerating collection processing. These monthly conversations are a core part of the Tier 3 charter program, which pairs early access and hands-on support with a community learning approach focused on responsible AI adoption. Through monthly meetings and structured working groups, participants are helping to shape both their own practices and the future direction of JSTOR’s digital stewardship tools.

During our first gathering, some charter participants shared plans to process items critical to institutional memory. Others are diving into political papers with potential global research value. Many are experimenting with new workflows, integrations, and long-missing capabilities, as well as taking on longstanding backlogs with new momentum.

Though contexts vary, a shared theme emerged: JSTOR Seeklight is expanding what’s possible without requiring institutions to compromise on what matters.

Tackling backlogs and fast-tracking access

The idea for JSTOR Seeklight emerged from listening closely to the field. Through hundreds of conversations with library leaders, site visits, and research interviews in 2024, one message was clear: the backlog is often overwhelming.

A cluttered metal shelf filled with stacks of old papers and folders. A blue sticky note taped to the shelf reads “SHELVES OF WOE, CONT’D” in black marker.

Processing deferred: Shelves of Woe

As born-digital records, donations, and digitized materials grow, capacity hasn’t kept pace. Minimal, box-level description is often the default, making it difficult for researchers—especially those outside the reading room—to find or use what’s been collected. In her recent field report on archives and special collections, Bridging Capacity and Care, Emilie Hardman quotes one institution as referring to these backlogs as “‘shelves of woe’—complex or vast unprocessed holdings that hinder access and mission fulfillment.”

We developed JSTOR Seeklight in response to this, asking ourselves: Could AI reduce processing bottlenecks? Surface hidden patterns? Free up time for expert work?

The charter community is already showing us that the answer is yes.

At an institution stewarding major political archives, staff plan to use JSTOR Seeklight to bring newly acquired materials into public view more quickly. “Normally that would go straight into our backlog,” one staff member said. “But we see an opportunity here to bring it to the front, learn something new, and diversify the representation of political figures in our collections.”

Another librarian shared the challenge of managing collections heavy in handwritten material, especially with student workers who often can’t read cursive:

“JSTOR Seeklight lets us imagine better descriptions–faster, and with more depth—than we could achieve otherwise.”

A small college is testing item-level description for the first time, comparing JSTOR Seeklight’s output to a student-created transcription of a 1940s diary. A global policy institute is exploring faster processing for in-demand materials, including papers from board members in Congress.

This theme—scaling processing in service of mission without sacrificing context—is echoed across the cohort. In each case, backlogs are being reframed as opportunities to surface long-overlooked histories. As one participant put it: “It’s great to have a way to get some of this backlog out of the closet.”

From “boxes under the table” to cloud-based stewardship

Participants also reflected on how JSTOR Seeklight and the broader JSTOR Stewardship platform support longer-term goals to modernize digital collections infrastructure and ensure sustainable access.

Many described fragmented ecosystems: combinations of legacy systems, self-hosted sites, and patchwork workflows that make digital stewardship difficult to scale. One librarian from a research-intensive university put it plainly: their environment is “a mix,” including “a homegrown system… and boxes under the table.”

Against this backdrop, institutions see the charter program as an opportunity to unify their approaches, integrating JSTOR Seeklight alongside existing tools or laying the groundwork for future migration. Several participants also highlighted the value of JSTOR’s integrated preservation services through Portico—which are included in the charter offering—to close long-standing gaps in their infrastructure and ensure long-term access. As one librarian put it:

“We’re really excited to modernize our infrastructure and workflows with this cloud-based program.”

Importantly, this investment in infrastructure translates directly to institutional impact and mission alignment. At one private research university, JSTOR Stewardship aligns with efforts to increase the visibility of underrepresented voices and create student internships that build workforce readiness. Another large research university described how JSTOR Seeklight offers a path to bring thousands of under-described collections out of the reading room and into digital discovery.

Together, these reflections underscore a shared shift from patchwork systems and limited capacity to cohesive, mission-aligned infrastructure that supports institutional goals.

Responsible AI in the hands of practitioners

Participants also emphasized the importance of introducing AI thoughtfully, as well as the opportunity for the charter cohort to work through these changes together. “Sometimes there’s a visceral reaction from practitioners,” said one digital collections head at a large public university. “There’s a need to be thoughtful about introducing these types of tools, and I’d love to know how you’re bringing people on board—what your institutions are concerned about and how you’re addressing it.”

That’s why JSTOR Seeklight has been intentionally designed for institutional oversight. Its metadata suggestions are accompanied by confidence scores and transparent tagging indicating what has been generated by AI and what has been reviewed or edited by expert practitioners. Practitioners maintain control over what is accepted, edited, or discarded—supporting thoughtful, responsible adoption.

That’s also why we’re introducing JSTOR Seeklight not as a finished product, but as a tool being shaped by the community using it. Through monthly conversations, working groups, and structured research, charter participants are directly informing how Seeklight evolves, from refining its metadata generation capabilities to helping guide its expansion into a more comprehensive processing solution. Their input ensures the technology remains grounded in archival values, responsive to real-world challenges, and aligned with evolving standards.

One participant, formerly in product development and software engineering, noted how excited they were to help shape the tool’s evolution:

“We want to improve our metadata, but we also want to help the community by co-developing this tool.”

A path toward inclusive, impactful access

At its core, JSTOR Digital Stewardship Services reflects a broader movement toward mission-aligned stewardship, where access, preservation, and ethical technology go hand-in-hand.

Participants are using JSTOR Seeklight to bring visibility to overlooked voices, build student learning opportunities, and digitize materials that have long sat inaccessible. Whether describing political records, institutional histories, or community narratives, institutions are doing more without compromising care, ethics, or context. And, they’re doing so in the context of a community committed to learning together, with and from each other, and in the interest of the long-term good of libraries and archives.

Interested in partnering to build values-driven, practitioner-informed AI for archives? Learn more about JSTOR Seeklight and the integrated JSTOR Digital Stewardship Services platform.

About the author

Alex Houston is a Senior Marketing Manager at ITHAKA with over 15 years of experience supporting the academic community. With a background in the scholarly publishing ecosystem, graduate coursework in philosophy, and freelance archival experience, she leads communication strategy for initiatives like JSTOR Digital Stewardship Services and is proud to help advance ITHAKA’s mission to expand access to knowledge and education.