EXPOSED
The classroom hour — live conversation practice, error correction with cultural context, seminar discussion of literature — still needs a person, and students pay for the cohort and the mentor. But the surrounding work (drafting syllabi and lesson plans, building vocabulary and grammar drills, generating reading glosses, grading translation exercises and short compositions, writing feedback) is exactly what LLMs do well, and free machine translation has already gutted the instrumental case for enrolling in a language sequence. The near-term threat to this occupation is not an AI teaching Spanish 201 — it is deans closing under-enrolled language departments, which is already happening.
This fall is concentrated in 2020 and has not recovered since.
Median pay $69,990 → $79,350 -9.3% 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
-0.2%
Percentage only. The projection counts a different population from the 19,830 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -0.2% 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.
~1,900 openings a year on average, including replacing people who leave.
TeacherLecturerProfessorInstructorArabic TeacherFrench TeacherGerman TeacherHebrew TeacherChinese TeacherGreek ProfessorItalian TeacherRussian TeacherSpanish TeacherSwahili TeacherArabic ProfessorFrench ProfessorGerman ProfessorHebrew ProfessorLanguage TeacherSpanish LecturerArabic InstructorBilingual TeacherCollege ProfessorFrench Instructor
Holding it up: trust premium . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Roughly half your week resists automation — the Spanish 201 conversation hour, correcting a student's aspect error while explaining why a Mexican speaker would phrase it differently, running a seminar on Rulfo where the point is the argument between students — but the other half (syllabus revision, drill and quiz banks, vocabulary lists, glossing readings, marking short compositions and translation sets, and writing the paragraph of feedback at the bottom) is now first-draftable in minutes, which is what pins this at 9 rather than the 14+ of a role whose outputs can't be prompted.
Some physical or field component You are in a room with bodies — modelling pronunciation and mouth position, reading the confusion on a face mid-drill, chaperoning the study-abroad term or the conversation table over dinner — but nothing you do requires a truck, a lab bench, or a hard hat, and the pandemic proved the course can be delivered over Zoom at real cost rather than total failure, which is what keeps this at 7 instead of 13.
No licence, no signature requirement No state licence gates postsecondary language instruction: a department can staff Elementary French with an adjunct, a graduate TA, or a native-speaker lecturer holding an MA, and the PhD is a hiring credential your dean can waive, not a statute that requires a named human to sign off on your students' credit hours — hence 2 rather than the 11+ of professions where deregistration ends the career.
Meaningful discretion You make real calls with consequences — placement and proficiency-level decisions, whether a suspiciously fluent essay is DeepL and therefore an academic-integrity referral, what to cut from a canon under a 14-week cap, grades that affect scholarships — but you make them inside departmental rubrics, common finals, and an appeals process where the chair and the dean own the final ruling, which is a 9 and not the 15 of someone whose signature is the last one on the file.
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 (9/20 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 (13/20) is whether buyers specifically pay for a person. Judgment and accountability (9/20) 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 24 of this occupation's 40 points (60%).
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 50/100, still EXPOSED.
Task-mix shift: once drill construction, glossing and first-pass composition grading are handed to LLMs, the residual day is live conversational repair, literary interpretation in seminar, and dissertation/thesis direction — the tier machines do worst. This raises the score only if departments survive to retain the judgment tier rather than closing entirely.
Faculty ownership of AI-use adjudication: campus academic-integrity policies (already being rewritten at most US universities) that assign the language instructor as the fact-finder on whether submitted work was machine-translated, with authority to fail or refer. Language faculty are uniquely positioned as the only competent judges of translation provenance.
Institutional or accreditor rules that make live human-assessed oral proficiency the only accepted credential — e.g. ACTFL OPI (Oral Proficiency Interview) or CEFR-aligned oral exams required by state teacher-certification boards, seminaries, or federal programs (Boren, Fulbright, DLI/FSI ratings). If a degree audit requires an in-person tester rather than an AI-scored oral, the human hour becomes non-substitutable for the credential, not just the learning.
Court and immigration interpretation/translation certification tied to teaching faculty — state court interpreter boards and ATA certification create signature-and-liability roles, but this attaches to the translator credential, not the professorship. A route only where faculty also hold the certifying role.
The limit. Every lever here raises per-survivor scores while doing nothing about the enrollment collapse that actually ends jobs. A department cut to two lines can have a higher resistance score and 80% fewer workers; the register does not measure headcount.
| New York-Newark-Jersey City, NY-NJ | 1,460 | $98,600 +24% |
| Los Angeles-Long Beach-Anaheim, CA | 800 | $83,310 +5% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 650 | $79,510 +0% |
| Boston-Cambridge-Newton, MA-NH | 630 | $99,100 +25% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 500 | $82,890 +4% |
| Chicago-Naperville-Elgin, IL-IN | 490 | $80,940 +2% |
| Portland-Vancouver-Hillsboro, OR-WA | 400 | $90,160 +14% |
| Dallas-Fort Worth-Arlington, TX | 330 | $78,480 -1% |
| Riverside-San Bernardino-Ontario, CA | 130 | $171,660 +116% |
| Fresno, CA | 70 | $167,160 +111% |
| Sacramento-Roseville-Folsom, CA | 130 | $125,980 +59% |
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 40. 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.