COOKED
The core of the job — coding, sorting, indexing, scanning and retrieving records, purging per retention schedules — is exactly what document AI and records-management software already do at scale, and the digitization wave that eliminated most paper rooms is still running. What remains is physical: pulling boxes, handling medical or legal originals, chain-of-custody moves within an office, and that keeps the role alive as a shrinking hybrid clerk/office-support position rather than a standalone file job. No license, no sign-off, no client relationship insulates it.
Part 2020 shock, part continued decline in the years since.
Median pay $32,710 → $43,600 +6.6% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
-15.9% 84,300 → 70,900 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -15.9% by 2034. This is the case where the score and the forecast agree, and it is the one worth taking seriously.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~7,300 openings a year on average, including replacing people who leave.
ClerkFilerListerIndexerCut FilerMap ClerkPre CoderCard FilerFile ClerkFile KeeperIndex ClerkClaims ClerkClerk TypistFiling ClerkRecord ClerkComputer AideImaging ClerkInvoice CoderLine AssignerRecord KeeperBrand RecorderCut File ClerkDocument ClerkScanning Clerk
Holding it up: embodiment . Weakest point: liability shield .
Core tasks are already automatable Alphanumeric sorting, indexing by patient MRN or case number, keying metadata into an EDMS, and flagging files past a retention date are rule-based operations that OCR plus a records system executes without a person; the 4 rather than 0 reflects the residual work of reconciling mislabeled or damaged originals that nobody scanned correctly the first time.
Some physical or field component You pull banker's boxes off shelves, run the scanner feed, unjam it, deliver charts to a floor, and shred — real physical work, but it happens in a climate-controlled office or basement records room with predictable shelving, which is why this sits at 7 and not with the field trades in the teens.
No licence, no signature requirement There is no state license, no certification exam, no signature line on anything you touch; if a file goes missing the HIM director or the custodian of record answers for it, and the 1 rather than 0 only acknowledges that HIPAA and court-records rules do formally name you as a handler.
Executes defined procedures on defined inputs The retention schedule, the filing convention, and the release-of-information policy are all written down before you arrive; your discretion tops out at deciding which supervisor to ask when a document doesn't match any existing category.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (4/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (1/20) is whether the law requires a licensed human to sign. Trust premium (2/20) is whether buyers specifically pay for a person. Judgment and accountability (2/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 5 of this occupation's 16 points (31%).
Embodiment (7/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 28/100, still COOKED.
Backfile conversion and legacy-paper handling in domains where originals must be physically preserved — court exhibit rooms under state record-retention statutes, hospital legacy charts, land/deed registries, and litigation hold boxes. Wet-signature originals, sealed exhibits, and radiographic film cannot be shredded post-scan, so physical pull/refile/chain-of-custody work persists as the last defensible tier. If a state judiciary or medical board tightens original-document retention rather than permitting scan-and-destroy, this holds or rises modestly.
Certified evidence-handling roles: if courts extend chain-of-custody attestation requirements (already routine for evidence technicians and some ediscovery custodians under FRE 901/902(13)-(14) authentication practice) to records custodians generally, requiring a named human to attest that a produced record is the complete original set. A named custodian-of-record affidavit is already demanded in many civil productions; formalizing it as a credentialed role would attach a person to the signature.
Task-mix shift: once indexing and OCR classification are automated, what remains is exception handling — deciding whether a document falls under legal hold, whether a purge is defensible, whether a privileged item was mis-tagged. Where the file clerk role merges into records-management/RIM analyst work carrying destruction-authorization sign-off, the surviving position owns consequential calls. This depends on the employer retaining a human approver on the retention-destruction workflow rather than auto-purging.
If HIPAA, GLBA, or CJIS-style access rules — or an insurer's cyber policy — bar sending records to third-party AI/cloud processing, scanning and indexing stay in-house and human for sensitive corpora. CJIS security policy already constrains cloud handling of criminal justice data.
The limit. Even with every lever, this occupation stays low. The levers preserve a shrinking remnant inside records-management or court-clerk job families, not the standalone file clerk title; headcount decline continues regardless of score. No route to a trust premium exists — no buyer pays extra for a human to file.
| Los Angeles-Long Beach-Anaheim, CA | 3,930 | $48,450 +11% |
| New York-Newark-Jersey City, NY-NJ | 3,260 | $44,710 +3% |
| Dallas-Fort Worth-Arlington, TX | 3,170 | $42,850 -2% |
| Chicago-Naperville-Elgin, IL-IN | 2,640 | $58,320 +34% |
| Houston-Pasadena-The Woodlands, TX | 2,580 | $41,400 -5% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 1,920 | $45,760 +5% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 1,700 | $43,270 -1% |
| Atlanta-Sandy Springs-Roswell, GA | 1,640 | $46,700 +7% |
| San Jose-Sunnyvale-Santa Clara, CA | 540 | $60,840 +40% |
| Santa Cruz-Watsonville, CA | 60 | $60,530 +39% |
| Salinas, CA | 80 | $59,040 +35% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 16. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.