EXPOSED
A large slice of the EE workday — writing specs, running load and thermal calculations, generating and checking schematics, drafting test plans, writing firmware and MATLAB/Python analysis, producing compliance documentation — is exactly the structured text-and-math work AI now does at usable quality, and EDA vendors are embedding it directly into the tools. What holds is the bench and field half: bringing up a board that fails for reasons no model predicted, probing with a scope, commissioning switchgear on site, and owning the decision that a design is safe to energize or ship. The modal worker here is an electronics/power designer without a PE; the stamped-drawing shield applies mainly to the power and building-systems tier, and it is regulatory rather than technical.
Headcount grew steadily across the period.
Median pay $98,530 → $120,630 -2.1% 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
+7.2% 192,000 → 205,700 on the projections basis
Growing, and only partly exposed
The BLS expects +7.2% more of these jobs by 2034, and at 52/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
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.
~11,700 openings a year on average, including replacing people who leave.
EngineerTest EngineerCable EngineerField EngineerLaser EngineerMeter EngineerPlant EngineerDesign EngineerControl EngineerProject EngineerResults EngineerService EngineerTesting EngineerUtility EngineerCircuits EngineerDivision EngineerLighting EngineerAutomation EngineerElectrical DesignerElectrical EngineerProtection EngineerSmart Grid EngineerSubstation EngineerDistribution Engineer
Holding it up: judgment & accountability . Weakest point: embodiment .
Mixed — a routine tier and a judgment tier Schematic capture, load-flow and thermal calcs, timing analysis, firmware, and NEC/IEC compliance write-ups are now credibly co-authored by tools, but the irreducible half — bringing up a first article that draws 3A when the model said 300mA, chasing an EMI failure at the test house, tuning a control loop against a real plant — keeps this at 11 rather than the 6 or 7 you'd assign a pure calculation-and-documentation role.
Some physical or field component An 8 reflects that most EEs are at a workstation in Altium or Cadence most weeks, but periodically at a bench with a scope, thermal camera and soldering iron, or in a plant during commissioning and factory acceptance testing — real hands-on hours, in controlled labs rather than the uncontrolled roofs and trenches that earn 13+.
Certification preferred, not legally required The PE licence exists and genuinely binds the power, utility interconnection and building-systems tier who stamp one-lines and coordination studies, but the modal electronics or embedded EE ships product under UL/FCC/CE conformity with the company as the responsible party and no personal licence at all — that split lands at 10, not the 15 a uniformly stamped discipline would get.
Exists to be accountable for ambiguous calls A 14 is earned by the calls that have no procedure: derating a part below datasheet limits for a 15-year field life, declaring an arc-flash boundary, signing off that a lithium pack's protection scheme is adequate, or authorizing first energization of switchgear with people standing nearby — decisions where being wrong burns hardware, kills a program schedule, or kills someone.
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 (11/20 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 (10/20) is whether the law requires a licensed human to sign. Trust premium (9/20) is whether buyers specifically pay for a person. Judgment and accountability (14/20) 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 33 of this occupation's 52 points (63%).
Embodiment (8/20) 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.
No occupation passed every test: close enough to electrical engineers on skills and subject matter, at least 10 points more resistant, no big jump in training, no new licence, no pay cut, and not shrinking on its own. That happens for 223 of the 654 occupations here that aren't SAFE, and it is worth stating plainly rather than leaving the section off.
The usual reason is that exposure travels with the skill profile. The jobs most similar to yours tend to be exposed for the same reasons yours is, so the near neighbours don't clear the gap — and the ones that do are a different kind of work, not a transfer of what you already know. Read that as a limit of this method, not a verdict that you're stuck: it only compares whole occupations, and it cannot see specialisation, industry, or anything you'd bring that isn't in a federal skill survey.
Here is that claim on your own job rather than in the abstract. These are the three occupations closest to this one by skill and subject matter — the places the work would most naturally transfer — with what the register scores them:
That is the whole problem in three lines. The nearest work is not meaningfully safer, so there is no move here that trades a similar skill set for a better verdict. This is not us running out of ideas — it is what the neighbourhood looks like.
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 66/100, still EXPOSED.
Functional-safety and cyber regimes formalizing a named human 'responsible engineer' for hazard sign-off: ISO 26262/IEC 61508 safety-case attestation, FDA 510(k) electrical safety, and the EU Cyber Resilience Act's requirement for an identified responsible person for product security. If auditors require a named engineer to attest that AI-generated calculations and firmware were independently verified, the accountability sits with a person by rule rather than by custom.
Genuine two-tier structure here: if EDA-embedded generation absorbs schematic capture, load/thermal calc, and compliance doc drafting, the residual day is bench bring-up, EMC debug, signal-integrity anomalies, and root-cause on failures with no clean model. That residual tier is the part current AI cannot do at usable quality, and task_resistance rises as a share even as absolute automatable work grows.
State PE licensure boards narrowing or eliminating the 'industrial exemption' that lets unlicensed engineers do in-house design — NCEES model law already pushes this, and states like Texas and North Carolina periodically revisit it. If electrical design for utility interconnection, data center power, or EV charging infrastructure required a PE seal from the responsible designer rather than a reviewing consultant, the shield extends from the building-systems tier into the modal electronics/power design job. Also watch IEEE 1547 / utility interconnection rules and NFPA 70E arc-flash study requirements naming a licensed engineer as the signer of AI-produced coordination studies.
Task-mix shift toward commissioning and field diagnostics — substation and switchgear commissioning, grid interconnection witness testing, data center power-up — as design hours shrink. Utility and NERC witness-test requirements that a qualified engineer be physically present during energization would harden this.
The limit. No plausible route on trust_premium: buyers of electrical design purchase a working, certified product, not a human's touch, and no client-facing brand attaches to the engineer. Any premium that appears will be for the seal or the attestation, which belongs under liability_shield.
| Los Angeles-Long Beach-Anaheim, CA | 8,970 | $139,780 +16% |
| New York-Newark-Jersey City, NY-NJ | 6,290 | $128,340 +6% |
| Detroit-Warren-Dearborn, MI | 6,060 | $114,390 -5% |
| Houston-Pasadena-The Woodlands, TX | 5,770 | $131,070 +9% |
| Dallas-Fort Worth-Arlington, TX | 5,720 | $124,390 +3% |
| Seattle-Tacoma-Bellevue, WA | 4,930 | $140,210 +16% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 4,450 | $126,420 +5% |
| Chicago-Naperville-Elgin, IL-IN | 4,290 | $113,520 -6% |
| San Jose-Sunnyvale-Santa Clara, CA | 4,160 | $187,060 +55% |
| Slidell-Mandeville-Covington, LA | 50 | $167,010 +38% |
| San Francisco-Oakland-Fremont, CA | 2,470 | $166,030 +38% |
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 52. 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.