Legal Secretaries and Administrative Assistants
In short: today's best AI models could do about 59% 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 60% by the end of 2028 if progress keeps its long-run pace. The conversations, in-person and hands-on work, about 40% of the time, stays with people.
Perform secretarial duties using legal terminology, procedures, and documents. Prepare legal papers and correspondence, such as summonses, complaints, motions, and subpoenas. May also assist with legal research.
156,280 people do this job in the US. $55,570 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 on3 h
- A person checks AI's work20 h
- Desk work AI can't do yet1 h
- Stays with people16 h
Legal Secretaries and Administrative Assistants, now: of a 40-hour week, AI could take on 3 hours, 20 hours become checking its work, 1 hours are desk work it can't do yet, and 16 hours stay with people.
The largest pieces: Maintain case files and the library (about 7 h, a person checks); Prepare and process legal documents (about 5 h, a person checks); Collect records for attorneys (about 2 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.
- 31%
- 28%
- 10%
- 5%
- 25%
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 59% 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, 1%. With Claude Mythos Preview (early), measured in April 2026, 33%. Today's estimate: 59%.
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, 0.064% went to this job's tasks in May 2026, the 213th largest of the 447 jobs that showed up at all; 3 of its 14 tasks showed up. Within each of Anthropic's methods, never across a change: 0.0092% to 0.018% (released February and March 2025); 0.016% to 0.022% (August 2025 to February 2026); 0.068% to 0.064% (April to May 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | 0.0092% | 540 of 708 | 2 of 14 |
| Released March 2025 | 0.018% | 431 of 703 | 3 of 14 |
| The week of Aug 4, 2025 (released September 2025) | 0.016% | 398 of 655 | 4 of 14 |
| The week of Nov 13, 2025 (released January 2026) | 0.024% | 362 of 672 | 4 of 14 |
| The week of Feb 5, 2026 (released March 2026) | 0.022% | 394 of 677 | 5 of 14 |
| April 2026 (released June 2026) | 0.068% | 202 of 434 | 3 of 14 |
| May 2026 (released June 2026) | 0.064% | 213 of 447 | 3 of 14 |
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 25% of this job's working time as physical work. For 25% of the working time, a robot has been shown doing most of the task somewhere: 25% in a purpose-built cell. 0.25 of 3, the 405th highest of the 923 jobs.
- Not physical 75%
- In a purpose-built cell 25%
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
40% of the working time: the conversation, in-person and hands-on parts of the tasks.
- Maintain case files and the library, the hands-on work, 19%
- Collect records for attorneys, the conversation, 6%
- Attend meetings and take notes, the in-person work, 5%
- Make photocopies, the hands-on work, 4%
- Receive and place calls, the conversation, 3%
- Send legal correspondence, the hands-on work, 2%
What becomes checking the machine
52% 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. Another 8% is desk work that is small if wrong, which can be handed over with lighter checks.
- Maintain case files and the library, costly if wrong, 19%
- Prepare and process legal documents, costly if wrong, 12%
- Collect records for attorneys, costly if wrong, 6%
- Complete forms for clients, costly if wrong, 5%
- Search for relevant law and decisions, costly if wrong, 5%
- Attend meetings and take notes, costly if wrong, 2%
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.
- 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.
- Briefing a model and checking its drafts. About 31% 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.
- Running AI inside the business systems. About 28% of the time is work inside business systems, which a model can do only once it is connected: knowing the systems well enough to set that up and watch it.
- The hands-on craft. About 25% of the time is hands-on work, which no model here does; robots are rated separately on this page.
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 . 14 tasks from .
| Task | ||||
|---|---|---|---|---|
| Organize and maintain law libraries, documents, and case files.Keeping the paper files and library in order, and the electronic case files.Done hourly or more by most who do it. Robots: in a purpose-built cell, OPEX Falcon+ RED document scanner / digital mailroom (deployed). Anthropic's rating. | 39% | 2.5 h50 min to 6.7 h | Systems 50%, hands-on 50% | 48%Desk part half by now |
| Prepare, proofread, or process legal documents, such as summonses, subpoenas, complaints, appeals, motions, or pretrial agreements.Summonses, motions and appeals prepared, proofread and filed.Done daily by most who do it. | 12% | 1.3 h25 min to 3.3 h | Documents 80%, systems 20% | 99%Half by now |
| Assist attorneys in collecting information such as employment, medical, and other records.Requests written and calls made for employment and medical records.Done daily by most who do it. | 12% | 1.7 h25 min to 4.2 h | Documents 50%, conversation 50% | 48%Desk part half by now |
| Attend legal meetings, such as client interviews, hearings, or depositions, and take notes.Sitting in on interviews, hearings and depositions, taking notes.Done more than yearly by most who do it. | 7% | 2.1 h50 min to 5 h | In person 70%, documents 30% | 30%Desk part half by now |
| Complete various forms, such as accident reports, trial and courtroom requests, and applications for clients.Reports, requests and applications filled in.Done more than weekly by most who do it. | 5% | 38 min17 min to 1.7 h | Documents 60%, systems 40% | 100%Half by now |
| Review legal publications and perform database searches to identify laws and court decisions relevant to pending cases.Database searches and reading of legal publications.Done more than yearly by most who do it. | 5% | 2.5 h50 min to 5 h | Documents 100% | 99%Half by now |
| Make photocopies of correspondence, documents, and other printed matter.Copying documents.Done hourly or more by most who do it. Robots: in a purpose-built cell, Ripcord robotic digitization platform (sold). Anthropic's rating. | 4% | 13 min4 min to 38 min | Hands-on 100% | Stays with people |
| Submit articles and information from searches to attorneys for review and approval for use.Passing articles and search results to the attorney.Done daily by most who do it. | 3% | 50 min17 min to 1.7 h | Documents 100% | 100%Half by now |
| Mail, fax, or arrange for delivery of legal correspondence to clients, witnesses, and court officials.Mailing, faxing and arranging delivery of letters to clients, witnesses and the court.Done daily by most who do it. Robots: in a purpose-built cell, Quadient DS-1200 G5iQ folder inserter (sold). Anthropic's rating. | 3% | 17 min8 min to 38 min | Hands-on 60%, systems 40% | 40%Desk part half by now |
| Receive and place telephone calls.Live calls.Done hourly or more by most who do it. | 3% | 8 min4 min to 25 min | Conversation 100% | Stays with people |
| Draft and type office memos.Memos written.Done more than monthly by most who do it. | 2% | 21 min8 min to 50 min | Documents 100% | 100%Half by now |
| Schedule and make appointments.Bookings in the calendar, with calls.Done several times a day by most who do it. AI observed on this task. | 2% | 13 min4 min to 38 min | Systems 60%, conversation 40% | 60%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.
- Paralegals and Legal Assistants74% now, 84% by 2028
- Secretaries and Administrative Assistants, Except Legal, Medical, and Executive59% now, 68% by 2028
- Court, Municipal, and License Clerks75% now, 86% by 2028
- Executive Secretaries and Executive Administrative Assistants64% now, 71% by 2028
- Court Reporters and Simultaneous Captioners73% now, 79% by 2028
- Correspondence Clerks85% now, 85% 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 Legal Secretaries and Administrative Assistants is within reach of AI now?
- An estimated 59% of the working time is within reach of AI models now and 60% by the end of 2028, on METR's long-run pace at four in five. 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 Legal Secretaries and Administrative 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: Maintain case files and the library (48% of the task's time now); Prepare and process legal documents (99% of the task's time now); and Collect records for attorneys (48% of the task's time now).
- Which tasks of Legal Secretaries and Administrative Assistants stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Make photocopies; and Receive and place calls. Together such tasks are about 7% of the working time, and the people and body parts of the other tasks stay with people too.
- What should Legal Secretaries and Administrative 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; handling protected records; and briefing a model and checking its drafts. By the end of 2028 on the long-run pace at four in five, about 40% of the working time stays with people, 52% becomes checking a model's work and 8% 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 Legal Secretaries and Administrative Assistants?
- In the Claude.ai conversations Anthropic sampled for May 2026 and tied to a work task, 0.064% went to this job's tasks, the 213th largest share of the 447 jobs that showed up; 3 of its 14 tasks appeared. Anthropic's labor market measure puts the share of the job's tasks seen in real Claude use at work at 0%. 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 Legal Secretaries and Administrative Assistants?
- Anthropic's robot exposure rating (September 2026) finds 25% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 25% of the working time. Its index is 0.25 of 3. One example: OPEX Falcon+ RED document scanner / digital mailroom for "maintain case files and the library". 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 14 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.