COOKED
The core of this job — collecting application data, pulling credit reports, verifying income and employment documents, checking files for completeness, and chasing applicants for missing paperwork — is exactly what document-extraction AI and loan origination systems already do end to end. There is no license requirement, the underwriting decision belongs to someone else, and the borrower-facing contact is increasingly a portal notification rather than a phone call. The surviving fraction is exception handling: messy self-employed income, fraud red flags, and applicants who need a person to walk them through a confusing file.
Most of this decline happened after 2021 — it is not the pandemic dip.
Median pay $40,640 → $50,020 -1.5% 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
-2.3% 177,600 → 173,500 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -2.3% 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.
~13,300 openings a year on average, including replacing people who leave.
CloserLoan ClerkLoan CloserPost CloserCredit ClerkInvestigatorLoan AnalystClosing AgentLoan AssistantLoan ExpeditorLoan InspectorLoan ProcessorLoan SecretaryMortgage ClerkLoan OriginatorMortgage BrokerLoan InterviewerDisbursement ClerkDocument ProcessorMortgage ProcessorClosing CoordinatorLoan RepresentativeDocument CoordinatorFinancial Specialist
Holding it up: trust premium . Weakest point: liability shield .
Core tasks are already automatable Keying 1003 fields, ordering credit and flood certs, stacking W-2s and bank statements against a checklist, and sending the same three 'we still need your 2023 Schedule C' emails is work that Encompass/nCino plus OCR income-calculation tools now complete without a clerk touching the file, which is why this sits at 3 rather than mid-band.
Fully desk- and screen-based The only physical acts are handling a wet-signature packet, running a scanner, and pulling a fax — a 2 rather than 0 because branch clerks still take walk-in borrowers' original pay stubs and IDs across a counter.
No licence, no signature requirement No NMLS license is required for clerical intake as long as you don't quote terms or negotiate, so nothing you certify carries your name; the 2 reflects only that some employers require a lender-specific compliance certification and BSA/ID-verification training you can be written up for skipping.
Executes defined procedures on defined inputs You decide whether a paystub is stale, whether a VOE needs a re-call, and when to escalate a red flag, but the credit decision, exception approval, and any condition waiver go to the underwriter or loan officer under investor guidelines you cannot override — narrow calls inside a written checklist, hence 4.
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 (3/20 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 (2/20) is whether the law requires a licensed human to sign. Trust premium (5/20) is whether buyers specifically pay for a person. Judgment and accountability (4/20) 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 11 of this occupation's 16 points (69%).
Embodiment (2/20) 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.
Loan Officers EXPOSED
Accountants and Auditors EXPOSED
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 35/100 — EXPOSED.
Task-mix shift: if automated origination absorbs the full clean-file tier (W-2 income, verified via Plaid/The Work Number, e-close), what remains staffed is exception work — self-employed and gig income reconstruction from bank statements and Schedule Cs, synthetic-identity and document-tampering red flags, and borrowers whose files fail automated verification. This is a genuine two-tier job, but the residual tier is small and headcount shrinks with it: the score rises while the occupation contracts.
A named-human-reviewer requirement for adverse action on AI-scored applications. Watch: CFPB Circular 2023-03 already says lenders must give specific ECOA reasons even from black-box models; Colorado SB 24-205 (effective 2026, consequential decisions including lending) requires an opportunity to appeal to human review; a Reg B amendment or state analogue that names an identified employee who documents and signs the second-look denial would convert clerk work into a compliance-attested function.
Formalizing the fraud-referral call: if BSA/AML expectations or GSE repurchase-risk practice push lenders to require a named human to escalate or clear misrepresentation flags before submission (rather than a model score routing the file), the exception handler owns a consequential, second-guessable call rather than moving paper.
CFPB Section 1071 small-business lending data rule (compliance dates now 2026-2027) creating an attested-accuracy role for demographic and application-data collection, where a specific employee certifies the submitted data set. Certification duties are what turn clerical capture into a liability-bearing seat.
Narrow and channel-specific: credit unions, CDFIs, and reverse-mortgage/HECM origination, where HUD already mandates third-party counseling with a live counselor, and where members choose the institution partly for a person to walk the file. Any extension of a live-human-contact mandate to other high-risk products (e.g. state rules on manufactured-home or high-cost lending) would broaden it.
The limit. Even with every lever, this stays low-scoring. The underwriting decision belongs to someone else by design, there is no license to attach personal liability to, and any human-review mandate can be satisfied by a handful of reviewers per thousand files — raising the score of a much smaller occupation. Trust premium above the credit-union/HECM niche has no realistic route; most borrowers prefer the portal.
| Dallas-Fort Worth-Arlington, TX | 6,300 | $51,470 +3% |
| Los Angeles-Long Beach-Anaheim, CA | 5,410 | $54,430 +9% |
| Detroit-Warren-Dearborn, MI | 4,910 | $51,500 +3% |
| New York-Newark-Jersey City, NY-NJ | 4,710 | $60,190 +20% |
| Chicago-Naperville-Elgin, IL-IN | 4,560 | $53,870 +8% |
| Phoenix-Mesa-Chandler, AZ | 3,930 | $50,970 +2% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 2,790 | $53,780 +8% |
| Houston-Pasadena-The Woodlands, TX | 2,460 | $49,220 -2% |
| Salinas, CA | 60 | $65,160 +30% |
| San Jose-Sunnyvale-Santa Clara, CA | 820 | $62,530 +25% |
| Vallejo, CA | 100 | $62,160 +24% |
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.