AI reads handwritten historical documents and generates metadata from scanned collections faster than any human cataloguer. Evaluating which records have permanent historical value, and making them genuinely accessible to researchers, still requires professional training. Here is what the research says about the archivist profession in 2026, and what you can do about it.
Get My Personalised Fossil ScoreFossil Score
65
Species
Archaeopteryx
AI reads handwritten historical documents and generates metadata from scanned collections faster than any human cataloguer. Evaluating which records have permanent historical value, and making them genuinely accessible to researchers, still requires professional training.
Task Automation Risk
34%
of current archivist tasks are automatable with existing AI tools
Archivists select, preserve, and provide access to records of lasting historical significance — government documents, institutional records, photographs, film, correspondence, and manuscripts. AI is changing several core workflows: Transkribus and similar handwriting recognition AI can transcribe historical handwritten documents at scale, a task that previously took archivists hundreds of hours per collection. Large language models can generate finding aid descriptions from inventory lists and catalogue metadata automatically. Digitisation programmes at the Library of Congress, the National Archives, and major university libraries are using AI to process backlogs that would take decades of manual effort. What this does not replace is the archival appraisal judgment — the decision about which records from an organisation's operations have sufficient historical, evidential, or informational value to warrant permanent preservation. That judgment requires understanding institutional history, records management practice, and historical significance in ways that AI cannot reliably replicate. Reference work — helping a researcher identify which collections might contain what they're looking for — draws on contextual knowledge that goes well beyond metadata search. Digital preservation management (format migration, integrity monitoring, ingest workflows) is a growing technical specialisation that AI tools assist but do not eliminate. The profession is stable: federal, state, and university archival programmes maintain steady demand, and the volume of born-digital records requiring archival management is growing faster than staffing.
Task Autopsy
🦕 Class A — At Risk Now
🦅 Class C — Protected
Your AI Toolkit
You don't need to learn all of these. Pick one, use it for a week, and see how it fits into your work. Most have free options so you can try before you commit.
The open-source archival management system used at most US research universities and many state archives — competency in ArchivesSpace is expected for professional archivist positions at academic institutions
Try it ↗AI handwriting recognition platform for historical manuscripts — used by archives globally to transcribe large collections of handwritten documents; understanding how to train and validate its models is a growing skill
Try it ↗Digital collections management platform used at public and academic libraries — handles digitised photographs, documents, and manuscripts with AI-assisted metadata and full-text search
Try it ↗Draft finding aid descriptions, research archival standards, prepare grant narratives for digitisation projects, and explore digital preservation policy language
Try it ↗Process large volumes of inventory text into structured finding aid drafts, research archival theory and methodology, and write collection-level descriptions from box-level inventories
Try it ↗Digital preservation, metadata standards, and digital humanities courses — supports the technical specialisation in born-digital archives that is the highest-growth area of the profession
Try it ↗Extinction Timeline
Transkribus-based handwriting transcription and AI metadata generation are already in production use at major archives. Digitisation backlogs that previously faced decades of manual processing are being addressed with AI-assisted workflows. Professional appraisal and reference functions are unchanged.
By 2028, AI will handle first-pass description and metadata generation for most newly digitised collections, freeing archivists for complex appraisal, outreach, and reference work. Born-digital records management (managing email archives, electronic records systems) becomes the dominant technical challenge and growth area.
By 2031, archivists who can manage born-digital archival programmes — ingesting, preserving, and providing access to electronic records at scale — are the most in-demand professionals in the field. Physical manuscript work continues at specialised institutions. The appraisal and access functions that require professional judgment are unchanged.
For the transcription and initial metadata generation work, AI is already doing significant amounts of what used to be manual. For the judgment work — appraising which records merit permanent preservation, arranging complex collections, and helping researchers navigate holdings — no. Archival appraisal is a professional responsibility that requires contextual knowledge AI does not have. The Society of American Archivists certifies archivists specifically because these decisions require trained judgment.
Transkribus is an AI platform developed at the University of Innsbruck that uses handwriting recognition to transcribe historical manuscripts, letters, and documents. It can be trained on specific historical hands — 18th century German script, Victorian letter writing — and achieves accuracy rates that make it genuinely useful for mass transcription of large collections. This reduces the time required to make historical documents searchable from decades to months.
Digital preservation expertise — managing the long-term storage, format migration, and accessibility of born-digital records is the fastest-growing competency gap in the profession. Electronic records management knowledge (SharePoint, M-Files, Laserfiche) is increasingly expected as organisations manage their active records archivally. Encoded Archival Description (EAD) and metadata standards (Dublin Core, MODS, METS) remain core competencies. Understanding how to evaluate and validate AI transcription outputs rather than accepting them uncritically is increasingly relevant.
The Bureau of Labor Statistics projects steady growth in archivist positions through 2032. Federal archives, state archives, university special collections, and corporate archives all maintain consistent demand. The volume of born-digital records being created by governments and institutions has outrun the archival capacity to manage them — this represents both a professional challenge and a source of sustained employment.
Take the free Fossil Score assessment at DontGoDinosaur.com. It looks at your specific daily tasks — not just your job title — and gives you a personalised risk score, a breakdown of which tasks are most vulnerable, and practical steps you can take in the next 6 months. It takes about 4 minutes.
More in Education & Library
Forestry and Conservation Science Teachers
Forestry and Conservation Science Teachers are in a strong position. The core of this job — working with people, making judgment calls, solving unique problems — is hard for AI to touch.
Business Teachers
AI tutors handle content delivery, practice problems, and basic concept clarification at a level that was previously impossible to scale. The business professor running a live case study discussion where students must defend positions, mentoring a student through a career decision based on knowing their specific goals and limitations, or designing a course around a live consulting project with a real company is doing work AI cannot replicate.
Psychology Teachers
Psychology Teachers are in a strong position. The core of this job — working with people, making judgment calls, solving unique problems — is hard for AI to touch.
Social Sciences Teachers
Social Sciences Teachers are in a strong position. The core of this job — working with people, making judgment calls, solving unique problems — is hard for AI to touch.
Economics Teachers
Economics Teachers are in a strong position. The core of this job — working with people, making judgment calls, solving unique problems — is hard for AI to touch.
Advertising and Promotions Managers
AI is automating media buying, ad copy testing, and campaign optimisation. The strategic and creative leadership role survives, but the execution layer underneath it is shrinking fast.
Further reading
Your Personal Score
Get a Fossil Score built on your actual daily tasks, not a category average. 4 minutes. Free.
Calculate My Personal Fossil Score