EXPOSED
The modal worker here is a retail broker or financial services sales agent whose analytical work — screening securities, building model allocations, drafting client proposals, summarizing research, generating suitability documentation — is exactly what robo-advisors and LLMs already do at usable quality and near-zero marginal cost. What survives is the licensed, relationship-carrying part: FINRA Series 7/63/65 registration, Reg BI suitability accountability, and clients who pay for a named human to answer the phone when markets drop 20%. Institutional and complex-product desks (derivatives, structured products, capital markets sales) hold up better than commission-driven retail product sales, which robo platforms have been compressing for a decade.
Headcount grew steadily across the period.
Median pay $62,270 → $78,660 +1.1% 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
+3.3% 514,500 → 531,600 on the projections basis
Growing, and only partly exposed
The BLS expects +3.3% more of these jobs by 2034, and at 51/100 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.
~38,100 openings a year on average, including replacing people who leave.
BuyerBankerBrokerDealerTraderDay TraderSpeculatorBond BrokerBond TraderFlow TraderLoan BrokerBlock TraderFloor BrokerFloor TraderGrain BrokerMarket MakerStock BrokerStock TraderBranch BankerEnergy TraderEquity TraderMoney ManagerOnline TraderSales Advisor
Holding it up: trust premium . Weakest point: embodiment .
Mixed — a routine tier and a judgment tier Screening tickers, rebalancing to a model, running Monte Carlo retirement projections and drafting the pitch book are all deliverables Betterment and a spreadsheet already produce, which pins this at 9 rather than 14 — the residual non-automatable work is the discovery conversation, the objection handling, and the cold-call-to-funded-account conversion, not the analysis.
Fully desk- and screen-based Everything from order entry to CRM notes to Zoom client reviews happens on two monitors; the 3 rather than 0 reflects that book-building still runs partly on in-person seminars, golf, and branch walk-ins that a purely remote agent loses.
Licensed human required and personally liable Series 7 plus 63/65 or 66 is a hard gate — an unregistered person cannot solicit a trade, and Reg BI puts the recommendation on the named registered rep with FINRA arbitration and U4/U5 disclosure attached — but the 12 rather than 18 is because the broker-dealer's supervisory system and E&O absorb most claims, and firms substitute registered bodies routinely.
Meaningful discretion Calls on whether an illiquid alt or an annuity is suitable for a 68-year-old with concentrated employer stock are genuinely contestable and defended in arbitration, but the 12 rather than 16 reflects that compliance-approved product menus, model portfolios, and pre-trade suitability screens bound the discretion before the rep exercises it.
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 (12/20) is whether the law requires a licensed human to sign. Trust premium (15/20) is whether buyers specifically pay for a person. Judgment and accountability (12/20) 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 39 of this occupation's 51 points (76%).
Embodiment (3/20) 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.
Real Estate Brokers EXPOSED
Gambling Managers 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 65/100, still EXPOSED.
Task-mix shift: this occupation genuinely has two tiers. If screening, proposal drafting, and suitability documentation are fully absorbed, what remains is complex-product structuring, negotiated institutional execution, and crisis-window client persuasion — measurable as a shift in headcount toward capital-markets and structured-product desks and away from commission retail product sales
State-level fiduciary or best-interest annuity rules (NAIC model adopted in 45+ states) plus proposals to require a licensed producer of record for any AI-assisted annuity or insurance-wrapped product sale, keeping a personally liable human in every transaction chain
Continued growth in fee-based advisory relationships where clients pay explicit basis points for a named human of record, plus custodian/insurer requirements that a human be identified for account authorization and fraud-verification calls; visible in wirehouse and RIA pricing surviving alongside 25bp robo options
SEC/FINRA extending Reg BI so that an AI-generated recommendation delivered to a retail client requires a named registered representative to review and attest to suitability, with that rep personally exposed in arbitration — FINRA's 2024-25 guidance on supervision of generative AI tools and its Rule 3110 supervisory-control framework already point at named-person accountability for AI outputs; a rule that no 'recommendation' can be made by an algorithm without a registered principal's attestation would harden this
If FINRA arbitration and SEC enforcement continue naming the individual rep rather than the firm for AI-assisted recommendations, the role's remaining function becomes owning the call under ambiguity — concentration exceptions, illiquid alternatives, elderly-client capacity determinations under FINRA Rule 2165 senior-exploitation holds, which require a human judgment on file
The limit. Retail commission product sales has no strong route up on any dimension — robo compression has run for a decade and the trust premium there is thin and price-tested. The upside levers concentrate in institutional, complex-product, and high-net-worth segments, which are a minority of the 489k headcount, so occupation-level gains are capped in the mid-60s even if every lever fires.
| New York-Newark-Jersey City, NY-NJ | 59,450 | $167,890 +113% |
| Chicago-Naperville-Elgin, IL-IN | 24,080 | $100,760 +28% |
| Los Angeles-Long Beach-Anaheim, CA | 21,050 | $78,160 -1% |
| Dallas-Fort Worth-Arlington, TX | 15,370 | $86,740 +10% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 14,190 | $63,790 -19% |
| Phoenix-Mesa-Chandler, AZ | 12,220 | $77,780 -1% |
| Boston-Cambridge-Newton, MA-NH | 11,860 | $101,200 +29% |
| San Francisco-Oakland-Fremont, CA | 10,420 | $102,290 +30% |
| New York-Newark-Jersey City, NY-NJ | 59,450 | $167,890 +113% |
| Bridgeport-Stamford-Danbury, CT | 2,910 | $133,530 +70% |
| Elmira, NY | 50 | $116,910 +49% |
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 51. 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.