SAFE verdict contested
A large share of a lawyer's billable hours — memo drafting, contract markup, discovery review, precedent research, first-pass due diligence — is exactly what current models do at usable quality, and that leverage is already compressing associate-level work. What holds is the part only a licensed member of the bar can do: sign filings, appear in court, take depositions, negotiate live, and carry malpractice and ethical liability for the advice. The modal lawyer here is not a Big Law partner but a small-firm or in-house generalist whose value sits in client counsel and accountability, not document production.
Headcount grew steadily across the period.
Median pay $122,960 → $159,670 +3.9% 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
+4.1% 864,800 → 900,700 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +4.1% more of these jobs by 2034. Note that safe does not mean well paid — several of the fastest-growing resistant occupations are among the lowest paid on the register.
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.
~31,500 openings a year on average, including replacing people who leave.
LawyerCounselAttorneyBarristerSolicitorLaw WriterProsecutorTax LawyerConveyancerCivil LawyerTax AttorneyTitle LawyerTrial LawyerChief CounselCity AttorneyFamily LawyerLegal AdvisorLegal CounselPatent LawyerSports LawyerTown AttorneyCity SolicitorClaim AttorneyCounty Counsel
Holding it up: liability shield . Weakest point: embodiment .
Mixed — a routine tier and a judgment tier Document review, research memos, and first drafts of pleadings and contracts — the bulk of hours for anyone under partner — are now machine-assisted, but taking a deposition, examining a hostile witness, reading a judge in a motion hearing, and negotiating a settlement across the table stay stubbornly manual, which is why this lands at 10 rather than 5 or 15.
Some physical or field component Courthouse appearances, client site visits, jail interviews, deposition rooms, evidence inspections, and closings put lawyers off the screen on a regular cadence, but the venues are controlled and much litigation practice has already moved to Zoom hearings and e-filing, so this sits at 6 rather than in the trades range.
Licensed human required and personally liable Bar admission is jurisdiction-by-jurisdiction, Rule 11 makes the signing attorney personally sanctionable for filings, Model Rules 1.1 and 5.5 mean unauthorized practice is a criminal exposure and bad advice is a malpractice suit against a named individual — 18 not 20 only because in-house and staff roles diffuse some of the personal exposure.
Exists to be accountable for ambiguous calls Whether to advise a plea, whether a disclosure triggers a duty to a third party, whether to withdraw over a client's intended fraud, how much risk a merger rep carries — these are underdetermined calls with prison, insolvency, or disbarment on the other side, and the lawyer owns them alone.
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 (10/20 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 (18/20) is whether the law requires a licensed human to sign. Trust premium (16/20) is whether buyers specifically pay for a person. Judgment and accountability (17/20) 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 51 of this occupation's 67 points (76%).
Embodiment (6/20) 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 75/100, still SAFE.
Task-mix shift: this occupation genuinely has two tiers. If first-pass discovery review, memo drafting, and precedent research are absorbed, the residual day is deposition strategy, live negotiation, judgment on settle-vs-try, and client counseling under ambiguity — work that scores far higher. The observable marker is firms shrinking first/second-year associate classes while lateral demand for 5+ year litigators and deal leads holds.
State bars and courts formalizing what is now ad hoc: standing orders on AI disclosure/certification in federal district courts (post-Mata v. Avianca), plus ABA Formal Opinion 512 duties, hardening into rules that a licensed attorney must personally verify and certify every AI-assisted filing and client deliverable, with Rule 11 sanctions attaching personally. Also UPL enforcement against AI legal-advice products — e.g., the DoNotPay FTC action and state UPL suits — being extended to bar non-lawyer-owned AI counsel; conversely, Arizona-style ABS liberalization and Utah's sandbox cut the other way.
Malpractice insurers (e.g., ALAS, state bar mutuals) adding underwriting questions or premium differentials requiring documented attorney review of AI work product, and sophisticated clients writing 'named partner attention' or AI-use-disclosure clauses into outside counsel guidelines — a shift already visible in some corporate legal departments' OCGs. This converts human attention into a contractually priced item rather than a default.
Attorney-client privilege and work-product doctrine being held by courts not to extend to unsupervised AI-tool interactions, making the attorney's own judgment the only privileged locus of decision — raising the consequence-ownership of the human call.
The limit. Liability shield and judgment accountability are already near ceiling; the realistic upside is a few points, and it is concentrated in litigation and counseling roles. Document-production-heavy practice (doc review, small-firm form work, high-volume immigration or collections) has no route up on any dimension and is where the losses land regardless of what the bar does.
| New York-Newark-Jersey City, NY-NJ | 94,610 | $208,880 +31% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 46,840 | $195,190 +22% |
| Los Angeles-Long Beach-Anaheim, CA | 41,580 | $194,410 +22% |
| Chicago-Naperville-Elgin, IL-IN | 28,320 | $165,660 +4% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 23,270 | $160,320 +0% |
| Boston-Cambridge-Newton, MA-NH | 20,080 | $183,350 +15% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 19,040 | $170,660 +7% |
| Atlanta-Sandy Springs-Roswell, GA | 18,760 | $138,210 -13% |
| San Jose-Sunnyvale-Santa Clara, CA | 6,040 | $301,320 +89% |
| San Francisco-Oakland-Fremont, CA | 18,470 | $212,500 +33% |
| New York-Newark-Jersey City, NY-NJ | 94,610 | $208,880 +31% |
Mike Morse Law Firm · Morgan & Morgan · McCann FitzGerald · Dentons
NPR reports that courts in China have ruled in favour of workers dismissed due to AI-driven automation, amid continued job displacement anxiety.
Long Island Business News reports a law firm with Long Island offices has launched an AI initiative during its anniversary year.
Law firm Mintz reports that Connecticut has enacted legislation regulating employers' use of AI, including in employment decisions.
JD Supra reports that Connecticut has enacted an AI law imposing compliance obligations on employers, including around AI use in employment decisions.
A law firm briefing reports that Connecticut has enacted legislation regulating workplace AI use, requiring notice for certain AI applications and disclosure when reductions in force involve AI.
A legal blog reports Connecticut has enacted a new AI law with compliance obligations for employers, including in employment decision-making contexts.
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.