Skip to sources
Time Machine

Lawyers

Scrub through 335years of this role's history, from when it first emerged, through every wave of technology that reshaped it, to the cited projections for where it's heading next.

2026drag to travel through time
172517501775180018251850187519001925195019752000now
Country
2026
Known today as Lawyers (BLS SOC 23-1011; subspecializations tracked by ABA survey but not BLS)
US Employment
755K
OEWS is a point-in-time survey snapshot, not a continuous time series; BLS advises against using it for year-over-year trend comparison.
Median Annual Wage
$159,670
≈ $155,576 in 2024 dollars
Each dot is a cited figure over time; the dotted line only links them (values between aren't measured). Hollow dots are estimates.
Tools of the era

The tools that defined the work

Select an era to see how it reshaped the work.

  • Handwritten pleadings, quill pen, law library (Coke, Blackstone)

    The earliest American lawyers composed every document by hand. A brief in a significant case might run dozens of folio pages in the lawyer's own script; a contract was a single handwritten instrument, engrossed on parchment for durability. The law library was the only research tool: William Blackstone's Commentaries on the Laws of England (1765-1769) was the canonical text that shaped an entire generation of American lawyers, including Lincoln, who read it by firelight. The volume of litigation a single lawyer could handle was physically constrained by how fast he could write. The clerk who engrossed final copies was often a lawyer in training.

    Ledger workPaper recordkeeping
  • Typewriter + carbon paper; West Publishing National Reporter (1879)

    The Remington No. 1 typewriter reached commercial availability in 1874; by the late 1880s, law offices had adopted it for correspondence and short documents. More consequentially, John West launched the West Publishing Company's National Reporter System in 1879, creating the first comprehensive case-law compilation organized by region and indexed by subject. West's Key Number system — still in use today — gave lawyers a common vocabulary for finding precedent across the entire United States. By 1900, a competent attorney had access through West volumes to every significant appellate decision in the country. Carbon paper multiplied the legal secretary's output; a single typing produced up to six usable copies. The typewriter and West's reporters together tripled what a single practitioner could produce and cite in a given day.

    Work toolChanging equipment
  • IBM Selectric typewriter + dictation; photocopier (Xerox 914, 1959)

    The IBM Selectric (1961), with its rotating type-ball element, became the dominant law-office machine for two decades. It could produce clean, professional documents at twice the speed of earlier typewriters, and the ability to change type-balls allowed legal formatting (different point sizes, special symbols) without changing machines. The Xerox 914 plain-paper copier, introduced in 1959 and widely adopted through the 1960s, ended carbon-paper duplication and made discovery a volume business: for the first time, entire files of documents could be reproduced inexpensively for litigation. The photocopier is arguably the tool that created modern discovery practice — and, unintentionally, the need for paralegals to manage the resulting paper.

    Accounting softwareIntegrated ledgers
  • WordPerfect (law-office standard) and early practice management software

    WordPerfect 4.0 (1984) became the word processor of the legal profession for a decade, its macros for pleading paper formatting, line-numbering, and citation styles making it the default in firms of every size. The transition from Selectric to PC word processors meant lawyers and their secretaries could revise documents without retyping from scratch; it also meant that the typing pool — a significant overhead cost in large firms — shrank dramatically. Time-and-billing software (TABS, PCLaw) emerged in the mid-1980s as law firms began tracking matter profitability systematically.

    Work toolChanging equipment
  • Email + internet; federal CM/ECF e-filing (1996); e-discovery begins

    Email transformed legal correspondence from a multi-day process (letter, dictation, typing, delivery) to an instantaneous one — with the side effect of creating a discovery universe that grew exponentially every year. Federal courts began piloting the Case Management/Electronic Case Files (CM/ECF) system in 1996; by 2004, e-filing was mandatory in most federal districts. The PACER public-access system made federal case dockets freely searchable online. The explosion of email as a business communication tool simultaneously created the modern e-discovery industry: by 2000, a significant commercial litigation matter might involve millions of emails, and the question of how to review them had become the dominant cost driver in litigation.

    Work toolChanging equipment
  • DocuSign (2003) + Equivio predictive coding (2007)

    DocuSign was founded in 2003 and began displacing wet-signature closing processes in transactional law. By the early 2010s, it had become standard for routine commercial contracts, NDAs, and real estate closings — eliminating the closing-table ceremony for deals that didn't require one. Simultaneously, Equivio (founded 2004, acquired by Microsoft 2015) introduced supervised machine-learning for e-discovery document classification, giving litigators the first algorithmic tool for first-pass document review. Equivio's technology foreshadowed predictive coding, which in 2012 received its first court approval in Da Silva Moore v. Publicis Groupe.

    Work toolChanging equipment
  • Technology-Assisted Review (TAR) approved by courts; Kira and Luminance contract AI

    US Magistrate Judge Andrew J. Peck's ruling in Da Silva Moore v. Publicis Groupe (S.D.N.Y., Feb. 24, 2012) was the first federal court decision to explicitly approve predictive coding as a valid discovery methodology. The ruling opened the door for algorithmic review in the most document-intensive legal work. Kira Systems (2012) and Luminance (2015) brought machine learning to contract review — allowing lawyers to extract specific clause types from hundreds of contracts in minutes rather than days. By 2020, large-firm M&A due diligence was routinely running NDA and material-contract portfolios through contract AI before assigning associate review hours.

    Effect on the work

    Contract AI tools reduced the associate hours required for due diligence on large M&A transactions by an estimated 20-40% in the early 2020s; combined with LPO and offshore review, this reduced first-year BigLaw associate class sizes at several firms by 15-25% between 2017 and 2023.

    Work toolChanging equipment
  • Harvey AI (2022) + Allen & Overy firmwide GPT-4 deployment (Feb 2023)

    Harvey was co-founded in summer 2022 by Winston Weinberg (securities litigator) and Gabriel Pereyra (Google DeepMind). It raised $5M from the OpenAI Startup Fund in November 2022 and integrated GPT-4 at the model's launch. On February 16, 2023, Allen & Overy announced it had deployed Harvey firmwide to 3,500 lawyers, who had already run approximately 40,000 queries during a trial period. This was the first major BigLaw endorsement of a generative AI platform for substantive legal work — not document management or billing, but drafting briefs, researching case law, and summarizing contracts. The announcement triggered a competition among Am Law 100 firms that effectively made Harvey deployment a table-stakes signal of technological credibility by end of 2023.

    AI audit toolsPattern detection
  • Mata v. Avianca (June 2023): ChatGPT fabricates six case citations, federal sanctions follow

    On March 1, 2023, attorney Steven Schwartz of the New York firm Levidow, Levidow & Oberman filed a legal brief in Mata v. Avianca, Inc. (S.D.N.Y.) that cited six federal cases in support of his client's claims. None of them existed. Schwartz had used ChatGPT to research the brief and, when the AI produced confident-sounding case summaries complete with docket numbers, had filed them without verification. Opposing counsel flagged the citations; Schwartz doubled down by asking ChatGPT to confirm they were real, and ChatGPT confirmed they were real. On June 22, 2023, Judge P. Kevin Castel sanctioned Schwartz and his supervising partner $5,000 and ordered them to serve copies of the opinion on the judges whose names appeared in the fabricated citations. The case became the canonical example in every bar association AI ethics opinion filed in 2023-2024: the Magesh et al. (JELS 2025) finding of 17-43% hallucination rates across commercial legal AI platforms was, in part, a systematic study of the phenomenon Schwartz discovered the hard way.

    AI audit toolsPattern detection
  • Full BigLaw AI stack: CoCounsel, Lexis+ AI, Hebbia, Luminance, Relativity aiR, EvenUp

    By end of 2023 and accelerating through 2025-26, every major legal research platform had deployed a generative AI layer: Thomson Reuters acquired Casetext for $650M (June 2023) and its CoCounsel product reached one million users by February 2026; LexisNexis launched Lexis+ AI in October 2023 and upgraded to Lexis Protégé with 300+ pre-built legal workflows in February 2026; Harvey processed 10 million documents in Q1 2026 alone across its platform and was used by 80 of the Am Law 100; Relativity's aiR for Review (generally available Q3 2024) automated e-discovery classification with AI-generated reasoning. A&O Shearman's ContractMatrix (Harvey-powered) was used by 2,000 lawyers daily and saved 30% of contract review time. The Thomson Reuters Future of Professionals 2025 report projected AI would save each lawyer an average of 190 hours per year. The ABA's 2024 AI TechReport found 30% of all law firms were using AI tools, rising to 47.8% at firms with 500+ lawyers.

    Effect on the work

    Harvey M&A users reported up to 75% time savings on unstructured data rooms; A&O ContractMatrix saved 2-3 hours/week per lawyer; lateral associate hiring for AI-specialty positions grew 106% year-over-year (Law.com, March 2026). Am Law 100 first-year associate class sizes were down 15-25% from their 2022 peak by 2026.

    Work toolChanging equipment
Projection cone · present → 2034

What credible sources project

Scrub the slider past now to anchor each scenario on the scrubber. The spread is the range of futures credible sources project for this role.

Employment outlook
Projected change in the number of people doing this work.
Thomson Reuters Future of Professionals 2025
2030
+8%
Thomson Reuters' annual survey of legal professionals and firm leaders found that 80% of law firm leaders expect AI to fundamentally alter how they conduct business, while also projecting net demand growth for legal services as AI increases accessibility and lowers the cost of certain legal work, expanding the total addressable market. The +8% figure represents TR's net projection for the attorney employment pool: AI is expected to save each lawyer ~190 hours per year (4.7 weeks), which firms are projected to redeploy into new client work rather than reductions in headcount. The augmentation-expands-demand model is the most optimistic major projection in the cone and serves as the upper bound of the uncertainty range.
BLS Occupational Outlook Handbook 2024-34
2034
+4%
BLS Employment Projections — industry-occupation matrix + labor productivity assumptions. The 2024-34 OOH projects "faster than average" growth of 4% for lawyers (SOC 23-1011) from a 2023 base, with net openings of approximately 48,400 annually including separations and replacement needs. The BLS methodology explicitly accounts for AI augmentation but projects that demand for legal services will outpace productivity gains from AI, particularly driven by healthcare law, intellectual property, environmental regulation, and elder law for the aging population. This makes lawyers the mirror image of paralegals (which BLS projects flat or negative due to AI): AI is expected to substitute for paralegal tasks while expanding the reach and capacity of attorney practice.
AI task exposure
Share of the role’s tasks that researchers estimate AI can do. This is a measure of task exposure, not a forecast of jobs lost.
Goldman Sachs — legal task automation analysis (March 2023)
2030
15%
of tasks
Goldman Sachs' March 2023 analysis of generative AI's economic impact estimated that approximately 44% of legal work tasks are automatable by current generative AI — the highest sector-level exposure in the study, tied with administrative and office support work. The -15% employment projection is a curator-inferred application of the Goldman task-automation share to employment: Goldman projects that automation of 44% of tasks would translate to roughly 15% net employment reduction if one-third of automated tasks result in workforce reduction rather than redeployment to higher-value work. Goldman explicitly models the legal sector as having both high automation potential AND high demand growth — the net direction is contested. The Goldman report does not publish an occupation-level employment forecast for 23-1011 specifically.
Eloundou et al. — "GPTs are GPTs" (2023)
2024
8%
of tasks
GPT-4 task-by-task exposure labeling against O*NET task statements for legal occupations. Eloundou et al. found that legal occupations have among the highest GPT-4 'exposure' (β metric combining direct task capability and tool-assisted capability) of any major occupation group — a direct inversion of Frey & Osborne's low computerization probability. The difference reflects the two studies measuring different things: F&O measured 2013-era automation risk using rule-based algorithms; Eloundou measured 2023-era LLM capability on the actual tasks performed. Legal research, contract drafting, and regulatory analysis — the heart of associate-level law practice — score very high on Eloundou's β. The -8% figure represents a conservative translation of the high-exposure task share to a short-term employment effect, acknowledging that bar ethics rules, client expectations, and professional liability create adoption friction not present in the raw exposure metric.
Frey & Osborne (2013)
2033
3%
of tasks
Gaussian-process classifier on O*NET task features. Frey & Osborne rated lawyers at 0.035 probability of computerization — 35th lowest of 702 occupations, among the safest roles in their entire study. The rating reflects lawyers' tasks loading heavily on variables F&O coded as computerization bottlenecks: 'social intelligence' (reading adversarial and client dynamics), 'creativity' (constructing novel legal arguments), and 'negotiation' (real-time adjustment in settlement and deal contexts). This is the most optimistic projection in the cone; the -3% figure represents a researcher estimate of the employment effect of a 3.5% automation probability over 20 years at historical displacement rates. Note: F&O's paralegal rating (0.94) and lawyer rating (0.035) make this one of the cleanest task-vs-job distinctions in their entire dataset — the same information-handling work that makes paralegals highly exposed makes lawyers, who exercise judgment over that work, relatively protected.
Today, in this role

What's shifting in the work right now

The historical view above shows how this role has moved. This is the present-day detail: which AI tools are picking up which tasks, where the edge still is, and the natural directions this work can grow.

What's changing in your day

Three parts of your work where AI is already doing real lifting, and what stays yours.

AI is sitting alongside you hereOversee AI-driven M&A due diligence using Harvey or Hebbia Matrix: configure risk-extraction criteria for the target's data room, review AI-generated issue logs and contract summaries for material adverse change clauses, change-of-control provisions, and compliance exposures

Oversee AI-driven M&A due diligence using Harvey or Hebbia Matrix: configure risk-extraction criteria for the target's data room, review AI-generated issue logs and contract summaries for material adverse change clauses, change-of-control provisions, and compliance exposures; escalate high-risk findings and advise the deal team on negotiation strategy.[7],[12],[4]

Tools picking this up
Where your edge is

Learn to write precise diligence extraction schemas: define the specific clauses (MAC, CoC, exclusivity, non-compete) that matter for each transaction type. AI surfaces the issues; the attorney's value is interpreting their commercial significance and advising on deal structure. Develop industry-specific knowledge of what risk thresholds are market-standard vs. deal-breaking.

AI is sitting alongside you hereDirect AI legal research agents (CoCounsel Deep Research, Westlaw Advantage AI) to execute multi-jurisdictional case law surveys

Direct AI legal research agents (CoCounsel Deep Research, Westlaw Advantage AI) to execute multi-jurisdictional case law surveys; validate every AI-returned citation with Shepard's or KeyCite before citing in any brief or memo; synthesize findings into a research memorandum with jurisdiction-specific precedent hierarchy.[18],[10],[9]

Where your edge is

Master structured prompting for AI research workflows: specify jurisdiction, date range, binding vs. persuasive precedent, and negative treatment. Always Shepard's-check every AI-returned case — Magesh JELS 2025 found 17-33% hallucination rates even in retrieval-augmented legal AI tools. The lawyer who writes precise research instructions and validates outputs reliably is irreplaceable.

AI is sitting alongside you hereDesign and supervise AI-assisted eDiscovery workflows for litigation matters: configure responsiveness and privilege parameters in Relativity aiR

Design and supervise AI-assisted eDiscovery workflows for litigation matters: configure responsiveness and privilege parameters in Relativity aiR; QA AI classifications on a statistically defensible sample; certify the privilege log; and make final calls on borderline attorney-client and work-product privilege designations.[19],[3]

Where your edge is

Shift focus from document review volume to protocol design and privilege log defense. AI auto-classifies at scale; the attorney is responsible for the sampling strategy and the certifications submitted to opposing counsel and the court. Develop expertise in technology-assisted review (TAR/predictive coding) to be able to defend the process under Federal Rules of Civil Procedure Rule 26.

Where this role is heading

Natural next steps for someone with your foundation: not exits, evolutions.

A direction you could grow

Compliance Officers

Lawyers who practice regulatory, employment, or securities law possess deep compliance expertise that translates directly to in-house compliance officer roles — often without additional credentials. Corporate legal teams actively recruit JD-holding compliance professionals, particularly in highly regulated sectors (healthcare, financial services, tech). AI is accelerating the shift of outside-counsel work in-house, growing the compliance function even as it compresses law firm headcount. The pivot typically requires domain specialization in one regulatory framework (HIPAA, SOX, GDPR, SEC) but not retraining from scratch.

What you'd add
· Cross-functional stakeholder communication beyond legal department
What it takesMost of your skills carry over
Share this year
Drops anyone you send it to straight into 2026.
Preview card
Part of Legal · see all 6roles →
Different role?

See the same long-arc view for your own profession.

Browse the directory by industry, or search by title or SOC code. New roles ship every few weeks. Every profile cites every claim.

Browse all roles

The data behind this timeline

On record since1701
Latest tracked employment754,500 (US, 2025)
Latest median pay$159,670 (2025)
Outlook+4% by 2034 (BLS Occupational Outlook Handbook 2024-34)
View all 28 cited data points
YearUS employmentMedian annual paySource
1900114,500n/aCENSUS-DECENNIAL
1951221,605n/aESTIMATE
1970326,400n/aESTIMATE
1990777,000n/aESTIMATE
2000681,000$88,280BLS-OEWS
2003516,220$91,490BLS-OEWS
2004521,130$94,930BLS-OEWS
2005529,190$98,930BLS-OEWS
2006547,710$102,470BLS-OEWS
2007555,770$106,120BLS-OEWS
2008553,690$110,590BLS-OEWS
2009556,790$113,240BLS-OEWS
2010728,200$112,760BLS-OEWS
2011570,950$113,310BLS-OEWS
2012581,920$113,530BLS-OEWS
2013592,670$114,300BLS-OEWS
2014603,310$114,970BLS-OEWS
2015609,930$115,820BLS-OEWS
2016619,530$118,160BLS-OEWS
2017628,370$119,250BLS-OEWS
2018792,500$120,910BLS-OEWS
2019657,170$122,960BLS-OEWS
2020658,120$126,930BLS-OEWS
2021681,010$127,990BLS-OEWS
2022707,160$135,740BLS-OEWS
2023813,900$145,760BLS-OEWS
2024819,300$151,160BLS-OEWS
2025754,500$159,670BLS-OEWS
Embed this timeline on your site

Free for any site. Paste this where the timeline should appear; it stays interactive, every datapoint stays cited, and it sets no cookies on your page. How embedding works

<iframe src="https://futurehistory.earth/embed/23-1011"
  width="100%" height="430" style="border:0"
  title="Lawyers, a Future History timeline"
  loading="lazy"></iframe>

See all roles in Legal