Postal Service Mail Carriers
In short: today's best AI models could do about 4% 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 the same by the end of 2028 if progress keeps its long-run pace. The conversations, in-person and hands-on work, about 96% of the time, stays with people.
Sort and deliver mail for the United States Postal Service (USPS). Deliver mail on established route by vehicle or on foot. Includes postal service mail carriers employed by USPS contractors.
328,820 people do this job in the US. $60,550 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 on1 h
- A person checks AI's work0 h
- Desk work AI can't do yet0 h
- Stays with people39 h
Postal Service Mail Carriers, now: of a 40-hour week, AI could take on 1 hours, 0 hours become checking its work, 0 hours are desk work it can't do yet, and 39 hours stay with people.
The largest pieces: Keep delivery records (under an hour).
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.
- 0%
- 4%
- 0%
- 1%
- 95%
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 4% 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, 4%. Today's estimate: 4%.
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, no reportable share went to this job's tasks in May 2026; 0 of its 21 tasks showed up. Within each of Anthropic's methods, never across a change: 0.002% to 0.0045% (August 2025 to February 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | none | n/a | 0 of 21 |
| Released March 2025 | none | n/a | 0 of 21 |
| The week of Aug 4, 2025 (released September 2025) | 0.002% | 632 of 655 | 1 of 21 |
| The week of Nov 13, 2025 (released January 2026) | 0.0039% | 599 of 672 | 1 of 21 |
| The week of Feb 5, 2026 (released March 2026) | 0.0045% | 583 of 677 | 1 of 21 |
| April 2026 (released June 2026) | none | n/a | 0 of 21 |
| May 2026 (released June 2026) | none | n/a | 0 of 21 |
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 95% of this job's working time as physical work. For 94% of the working time, a robot has been shown doing most of the task somewhere: 77% in a purpose-built cell, 15% in a structured facility, 2% in any setting. 1.14 of 3, the 73rd highest of the 923 jobs.
- Not physical 5%
- No robot shown 1%
- In a purpose-built cell 77%
- In a structured facility 15%
- In any setting 2%
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
95% of the working time: the conversation, in-person and hands-on parts of the tasks.
- Deliver the route, the hands-on work, 63%
- Case mail in delivery order, the hands-on work, 15%
- Bundle mail for the route, the hands-on work, 4%
- Bring collected mail back, the hands-on work, 3%
- Keep to the collection schedule, the hands-on work, 2%
- Scan letters and parcels, the hands-on work, 2%
What becomes checking the machine
1% 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 3% is desk work that is small if wrong, which can be handed over with lighter checks.
- Register and insure mail, costly if wrong, 0.5%
What to learn
- The hands-on craft. About 95% of the time is hands-on work, which no model here does; robots are rated separately on this page.
- Judgment where errors are severe. Some tasks are severe if wrong. Knowing when not to trust an output matters more there than speed.
- 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.
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 . 21 tasks from .
| Task | ||||
|---|---|---|---|---|
| Deliver mail to residences and business establishments along specified routes by walking or driving, using a combination of satchels, carts, cars, and small trucks.Walking and driving the route with the mail.Done daily by most who do it. Robots: in a purpose-built cell, Berkshire Grey Robotic Put Wall (BG RPW) (sold). Anthropic's rating. | 63% | 3.8 h2.5 h to 5 h | Hands-on 100% | Stays with people |
| Sort mail for delivery, arranging it in delivery sequence.Sorting the route's mail by hand.Done daily by most who do it. Robots: in a purpose-built cell, USPS Delivery Point Sequencing on Delivery Barcode Sorters (DBCS) (deployed). Anthropic's rating. | 15% | 1.3 h25 min to 2.5 h | Hands-on 100% | Stays with people |
| Bundle mail in preparation for delivery or transportation to relay boxes.Bundling by hand.Done daily by most who do it. Robots: in a purpose-built cell, Felins banding/bundling systems (US-2000, M-series inline banders) (sold). Anthropic's rating. | 4% | 25 min13 min to 50 min | Hands-on 100% | Stays with people |
| Return to the post office with mail collected from homes, businesses, and public mailboxes.Driving the collected mail back to the office.Done daily by most who do it. Robots: in any setting, Cainiao GT / GT Pro L4 autonomous delivery vehicle (deployed). Anthropic's rating. | 3% | 21 min8 min to 38 min | Hands-on 100% | Stays with people |
| Maintain accurate records of deliveries.Entries on the scanner or the sheet.Done daily by most who do it. | 2% | 13 min4 min to 25 min | Systems 100% | 100%Half by now |
| Meet schedules for the collection and return of mail.Running the route to time.Done daily by most who do it. Robots: in a structured facility, Aethon T3 / TUG autonomous mobile robot (sold). Anthropic's rating. | 2% | 13 min4 min to 25 min | Hands-on 100% | Stays with people |
| Obtain signed receipts for registered, certified, and insured mail, collect associated charges, and complete any necessary paperwork.The signature at the door, the charge collected, and the paperwork done.Done daily by most who do it. Robots: in a structured facility, Aethon TUG (deployed). Anthropic's rating. | 2% | 10 min4 min to 21 min | Hands-on 50%, in person 30%, systems 20% | 20%Desk part half by now |
| Scan labels on letters or parcels to confirm receipt.Scanning each item as it goes.Done hourly or more by most who do it. Robots: in a structured facility, Spot last-mile delivery payload (Boston Dynamics) (demonstrated). Anthropic's rating. | 2% | 3 min1 min to 7 min | Hands-on 100% | Stays with people |
| Travel to post offices to pick up the mail for routes or pick up mail from postal relay boxes.Driving to the office or the relay box and loading up.Done daily by most who do it.Robots: in a structured facility, Digit (Agility Robotics) (deployed). Anthropic's rating. | 1.0% | 17 min8 min to 25 min | Hands-on 100% | Stays with people |
| Register, certify, and insure parcels and letters.Entering the item and labeling it.Done daily by most who do it.Robots: in a purpose-built cell, Advanced Facer Canceler System (AFCS / AFCS 200) (USPS) (deployed). Anthropic's rating. | 0.9% | 10 min4 min to 17 min | Systems 60%, hands-on 40% | 60%Desk part half by now |
| Turn in money and receipts collected along mail routes.Handing over the day's takings at the office.Done daily by most who do it. Robots: in a purpose-built cell, G+D robotic cash center cell (BPS M7 / second-level packaging) (Giesecke+Devrient) (deployed). Anthropic's rating. | 0.9% | 13 min4 min to 21 min | Hands-on 100% | Stays with people |
| Return incorrectly addressed mail to senders.Marking and returning the item.Done daily by most who do it. Robots: no robot shown. Anthropic's rating. | 0.8% | 4 min2 min to 13 min | Hands-on 100% | Stays with people |
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.
- Postal Service Clerks19% now, 19% by 2028
- Couriers and Messengers7% now, 7% by 2028
- Mail Clerks and Mail Machine Operators, Except Postal Service13% now, 13% by 2028
- Shipping, Receiving, and Inventory Clerks53% now, 53% by 2028
- Cargo and Freight Agents50% now, 50% by 2028
- Postal Service Mail Sorters, Processors, and Processing Machine Operators1% now, 1% 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 Postal Service Mail Carriers is within reach of AI now?
- An estimated 4% of the working time is within reach of AI models now and 4% 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 Postal Service Mail Carriers 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: Keep delivery records (100% of the task's time now); Register and insure mail (60% of the task's time now); and Get signatures for accountable mail (20% of the task's time now).
- Which tasks of Postal Service Mail Carriers stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Deliver the route; Case mail in delivery order; and Bundle mail for the route. Together such tasks are about 94% of the working time, and the people and body parts of the other tasks stay with people too.
- What should Postal Service Mail Carriers learn as AI takes on more of the work?
- By the job's own task-by-task review, the first things to learn are the hands-on craft; judgment where errors are severe; and reviewing work you did not do yourself. By the end of 2028 on the long-run pace at four in five, about 95% of the working time stays with people, 1% becomes checking a model's work and 3% 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 Postal Service Mail Carriers?
- In the Claude.ai conversations Anthropic sampled for May 2026, no reportable share went to this job's tasks. 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 Postal Service Mail Carriers?
- Anthropic's robot exposure rating (September 2026) finds 95% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 94% of the working time, 2% in any setting. Its index is 1.14 of 3. One example: Cainiao GT / GT Pro L4 autonomous delivery vehicle for "bring collected mail back". 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 21 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.