Paralegals and Legal Assistants
In short: today's best AI models could do about 74% of this job's working time (projected from METR's April 2026 measurement and Epoch's November 2026 index) if the work were set up for them, and about 84% by the end of 2028 if progress keeps its long-run pace. The conversations, in-person and hands-on work, about 16% of the time, stays with people.
Assist lawyers by investigating facts, preparing legal documents, or researching legal precedent. Conduct research to support a legal proceeding, to formulate a defense, or to initiate legal action.
392,880 people do this job in the US. $62,890 a year.
Reviewed on September 30, 2026: what each task is made of, and what handing it over would ask.
A typical 40-hour week
- AI could take on0 h
- A person checks AI's work29 h
- Desk work AI can't do yet4 h
- Stays with people7 h
Paralegals and Legal Assistants, now: of a 40-hour week, AI could take on 0 hours, 29 hours become checking its work, 4 hours are desk work it can't do yet, and 7 hours stay with people.
The largest pieces: Draft and review legal documents (about 9 h, a person checks); Research statutes and cases (about 6 h, a person checks); Draft legal documents and file them (about 5 h, a person checks).
Estimates for a typical week of this job, if progress keeps its long-run pace, counting work the best models finish four times in five. Every task is below.
of the working time. Now is projected from METR's April 2026 measurement and Epoch's November 2026 index.
- 77%
- 6%
- 7%
- 2%
- 8%
of the work: Anthropic's measure of what people already bring to Claude at work (2025). It counts help with part of a task, so it can be higher than the share within reach.
On the chosen pace and reliability. Estimate.
Share of the working time within reach, now to 2030
Change assumptionsHide assumptionsLong-run pace, four in five
What each choice means: , , , , . Every pace starts from .
At this date and pace, models finish tasks up to 2.4 work days long four times in five.
What handing it over would take, now
The 74% within reach splits by what a model would need. The dates, pace and reliability follow the controls above. Estimates.
of the working time. The material can be handed to a model and the result handed back.
of the working time. The model has to be connected to the system the work lives in.
is conversation short enough by length. This site leaves conversation with people, so it is not in the total.
How far this has come
With GPT-4, measured in March 2023, 0% of this job's working time was within reach at . With Claude 3.7 Sonnet, measured in February 2025, 0%. With Claude Mythos Preview (early), measured in April 2026, 27%. Today's estimate: 74%.
A backcast: today's task data with the best model METR had measured by each date. It shows how fast the models have moved, not what anyone forecast at the time.
Measured use over time
Of the Claude.ai conversations Anthropic sampled and tied to a work task, 0.16% went to this job's tasks in May 2026, the 120th largest of the 447 jobs that showed up at all; 2 of its 12 tasks showed up. Within each of Anthropic's methods, never across a change: 0.044% to 0.053% (released February and March 2025); 0.054% to 0.063% (August 2025 to February 2026); 0.15% to 0.16% (April to May 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | 0.044% | 296 of 708 | 3 of 12 |
| Released March 2025 | 0.053% | 262 of 703 | 2 of 12 |
| The week of Aug 4, 2025 (released September 2025) | 0.054% | 236 of 655 | 3 of 12 |
| The week of Nov 13, 2025 (released January 2026) | 0.066% | 201 of 672 | 3 of 12 |
| The week of Feb 5, 2026 (released March 2026) | 0.063% | 220 of 677 | 3 of 12 |
| April 2026 (released June 2026) | 0.15% | 128 of 434 | 2 of 12 |
| May 2026 (released June 2026) | 0.16% | 120 of 447 | 2 of 12 |
Measured, not estimated: Anthropic's samples of Claude.ai conversations, each tied by a model to an O*NET task, from the Anthropic Economic Index. It says where Claude's use goes, not how much of this job's working time it covers, so it is not on the same scale as the reach above. The releases differ in their sample, their classifier and the smallest share they report (the June 2026 file rounds each task to a hundredth of a percent, so a smaller task reads as none), which is why a change of method starts a new line. At least 96% of each older release's task-tied use matched this site's tasks, and all of the newest. Claude.ai only, so use through other tools and other companies' models is not in it. Every job ranked.
Robots and the physical work
Anthropic rates 1% of this job's working time as physical work. For 0% of the working time, a robot has been shown doing most of the task somewhere. 0.00 of 3.
- Not physical 99%
- No robot shown 1%
Anthropic's rating, not this site's estimate: Claude Opus 5, searching the web, rated each physical O*NET task by the kind of setting in which a robot it could cite (deployed, sold or demonstrated) does most of the task, weighted by its own estimate of each task's share of the time (What work can robots do?, September 30, 2026, data CC BY 4.0; every job ranked). A robot shown somewhere is not a robot in your building: cost, setup and safety decide that. The share within reach on this page still counts no body work. The robot behind each rated task is under the task below.
What changes for the people in this job
From the task-by-task review, by the end of 2028 on the long-run pace at four in five. It says which parts of the work stay with people and which change, not whether anyone keeps a job: that depends on decisions no data set can see.
What stays with people
16% of the working time: the conversation, in-person and hands-on parts of the tasks.
- Organize trial exhibits, the hands-on work, 6%
- Meet clients about their cases, the conversation, 4%
- Coordinate law office work, the conversation, 3%
- Assist closings and settle disputes, the conversation, 2%
- Inventory estate property, the hands-on work, 0.8%
What becomes checking the machine
83% of the working time: desk work a model could do by then, in tasks where a mistake costs something or a licensed person signs. A person checks it.
- Draft and review legal documents, severe if wrong, 27%
- Research statutes and cases, costly if wrong, 15%
- Investigate case facts and law, costly if wrong, 14%
- Draft legal documents and file them, costly if wrong, 13%
- Organize trial exhibits, severe if wrong, 6%
- Assist closings and settle disputes, severe if wrong, 3%
What to learn
- Reviewing work you did not do yourself. The tasks under What becomes checking the machine, where a mistake costs something, stay a person's to check: spotting what a draft gets wrong is a skill of its own.
- Briefing a model and checking its drafts. About 77% of the time is documents and data, the part models reach first: writing a clear brief, giving the right material, and checking what comes back.
- Handling protected records. Much of the work touches protected personal records: knowing what may be shared with a model, and how, is part of the job.
- Judgment where errors are severe. Some tasks are severe if wrong. Knowing when not to trust an output matters more there than speed.
Your own version of this job: read it against your job description. The headline is free; the full personal report, every task with its timeline and what to learn, is $19 (a sample).
The tasks
Sorted by . 12 tasks from .
| Task | ||||
|---|---|---|---|---|
| Prepare, edit, or review legal documents, including legislation, briefs, pleadings, appeals, wills, contracts, and real estate closing statements.Writing and checking briefs, contracts and wills.Done hourly or more by most who do it. | 27% | 2.5 h33 min to 1 work day | Documents 100% | 81%Half by now |
| Gather and analyze research data, such as statutes, decisions, and legal articles, codes, and documents.Reading and analyzing the law on the question.Done more than yearly by most who do it. AI observed on this task. | 15% | 2.5 h50 min to 6.7 h | Documents 100% | 96%Half by now |
| Investigate facts and law of cases and search pertinent sources, such as public records and internet sources, to determine causes of action and to prepare cases.Digging through records and sources to build the case.Done more than weekly by most who do it. | 14% | 3.3 h50 min to 1.3 work days | Documents 100% | 80%Half by now |
| Prepare affidavits or other documents, such as legal correspondence, and organize and maintain documents in paper or electronic filing system.Drafting affidavits and letters, and keeping the case files in order.Done daily by most who do it. | 13% | 1.1 h25 min to 2.5 h | Documents 60%, systems 40% | 100%Half by now |
| Prepare for trial by performing tasks such as organizing exhibits.Sorting and labeling the exhibits, on paper and in binders.Done yearly or less by most who do it. | 13% | 3.3 h50 min to 1.3 work days | Documents 50%, hands-on 50% | 40%Desk part half by now |
| Arbitrate disputes between parties and assist in the real estate closing process, such as by reviewing title searches.Reviewing titles and papers, and working the parties toward agreement.Done more than yearly by most who do it. | 6% | 2.1 h50 min to 5 h | Documents 60%, conversation 40% | 60%Desk part half by now |
| Direct and coordinate law office activity, including delivery of subpoenas.Directing the office's day, including who serves what.Done yearly or less by most who do it. | 5% | 1.3 h25 min to 3.3 h | Conversation 60%, documents 40% | 40%Desk part half by now |
| Meet with clients and other professionals to discuss details of cases.Meetings and calls with clients and other professionals.Done more than monthly by most who do it. | 4% | 50 min25 min to 2.1 h | Conversation 50%, in person 50% | Stays with people |
| Appraise and inventory real and personal property for estate planning.Listing and valuing the property, some of it seen in person.Done yearly or less by most who do it. | 2% | 3.3 h1.3 h to 1 work day | Documents 60%, hands-on 40% | 57%Desk part half by now |
| File pleadings with court clerks.Filing through the court's system, sometimes at the counter.Done more than monthly by most who do it. | 1% | 25 min13 min to 1.3 h | Systems 80%, hands-on 20% | 80%Desk part half by now |
| Keep and monitor legal volumes to ensure that the law library is up-to-date.Shelving updates and checking the subscriptions.Done yearly or less by most who do it.Robots: no robot shown. Anthropic's rating. | 0.1% | 50 min25 min to 2.5 h | Hands-on 60%, systems 40% | 40%Desk part half by now |
| Call upon witnesses to testify at hearings.Calls and letters to witnesses to get them to the hearing.Done yearly or less by most who do it. | 0.1% | 38 min17 min to 1.7 h | Conversation 70%, documents 30% | 30%Desk part half by now |
Within reach, per task: the share of the task's time whose instances are short enough for a model to finish at the chosen date, pace and reliability. Half by: the date half of that time comes within reach. Made of: how the task's time splits among five kinds of doing. Only the documents and systems parts can count as within reach, so a task that is 30% documents can be at most 30% within reach.
Jobs that share skills with this one
From O*NET's related occupations. Within reach now and by the end of 2028, long-run pace, four in five.
- Legal Secretaries and Administrative Assistants59% now, 60% by 2028
- Title Examiners, Abstractors, and Searchers92% now, 93% by 2028
- Court Reporters and Simultaneous Captioners73% now, 79% by 2028
- Court, Municipal, and License Clerks75% now, 86% by 2028
- Secretaries and Administrative Assistants, Except Legal, Medical, and Executive59% now, 68% by 2028
- Executive Secretaries and Executive Administrative Assistants64% now, 71% by 2028
Where this job works
The industry groups that employ the most of it (BLS, May 2025).
Where the most of these jobs are
The metro areas with the most people in it (BLS, May 2025).
Questions about this job
- How much of the work of Paralegals and Legal Assistants is within reach of AI now?
- An estimated 74% of the working time is within reach of AI models now and 84% by the end of 2028, on METR's long-run pace at four in five (83% to 84% across the three paces the site shows). Within reach means a model could finish work that long at that reliability. It is not a forecast of jobs lost or of savings.
- Which tasks of Paralegals and Legal Assistants come within reach first?
- Weighted by the time each takes, the tasks with the most within reach now (projected from METR's April 2026 measurement and Epoch's November 2026 index) are: Draft and review legal documents (81% of the task's time now, 100% by the end of 2028); Research statutes and cases (96% of the task's time now); and Draft legal documents and file them (100% of the task's time now).
- Which tasks of Paralegals and Legal Assistants stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Meet clients about their cases. Together such tasks are about 4% of the working time, and the people and body parts of the other tasks stay with people too.
- What should Paralegals and Legal Assistants learn as AI takes on more of the work?
- By the job's own task-by-task review, the first things to learn are reviewing work you did not do yourself; briefing a model and checking its drafts; and handling protected records. By the end of 2028 on the long-run pace at four in five, about 16% of the working time stays with people, 83% becomes checking a model's work and 0% could be handed over with lighter checks. It says which parts of the work change, never whether anyone keeps a job.
- Do people already use AI for the work of Paralegals and Legal Assistants?
- In the Claude.ai conversations Anthropic sampled for May 2026 and tied to a work task, 0.16% went to this job's tasks, the 120th largest share of the 447 jobs that showed up; 2 of its 12 tasks appeared. Anthropic's labor market measure puts the share of the job's tasks seen in real Claude use at work at 29%. These are measures of use, not of how much of the work AI could finish, and they cover Claude only.
- Can robots do the physical work of Paralegals and Legal Assistants?
- Anthropic's robot exposure rating (September 2026) finds 1% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 0% of the working time. Its index is 0.00 of 3. A robot shown somewhere is not one in every workplace; this site's own estimates still count no physical work as within reach.
- How is this estimated?
- From the 12 tasks O*NET lists for the occupation, each with an estimated length checked against GDPval and a share of the working time, reviewed one task at a time on September 30, 2026 for what each is made of (documents, business systems, conversation, in person, hands-on), and METR's measurements of how long a task the best AI models can finish, carried forward to the date from the latest, Claude Mythos Preview (early) in April 2026. Data checked November 10, 2026.
This job in your state or metro area
Embed these figures on your site
A small card with the share of working time within reach now and at the end of 2027, 2028 and 2030, the date and a link back. Free under CC BY 4.0; the card carries the credit.
Your own version of this job
The page above is the occupation. Your job is your title and what you actually do: give both, and they are read against these tasks in your browser, nothing sent anywhere. The headline is free: how much of your working time is within reach now and by the end of 2028, and how many of the tasks your description names.
The full personal report is $19, to print or save as a PDF: every task with when it comes within reach, which ones your description names, the parts of your description the task list does not cover, what stays with you, what becomes checking the machine and what to learn.