Tellers
In short: today's best AI models could do about 48% 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 52% of the time, stays with people.
Receive and pay out money. Keep records of money and negotiable instruments involved in a financial institution's various transactions.
329,480 people do this job in the US. $43,030 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 on2 h
- A person checks AI's work17 h
- Desk work AI can't do yet0 h
- Stays with people21 h
Tellers, now: of a 40-hour week, AI could take on 2 hours, 17 hours become checking its work, 0 hours are desk work it can't do yet, and 21 hours stay with people.
The largest pieces: Find balancing errors (about 3 h, a person checks); Process deposits and contributions (about 2 h, a person checks); Maintain loan records (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.
- 8%
- 40%
- 4%
- 9%
- 39%
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 48% 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, 14%. With Claude Mythos Preview (early), measured in April 2026, 48%. Today's estimate: 48%.
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, 0.011% went to this job's tasks in May 2026, the 410th largest of the 447 jobs that showed up at all; 1 of its 28 tasks showed up. Within each of Anthropic's methods, never across a change: 0.011% to 0.012% (released February and March 2025); 0.0017% to 0.011% (August 2025 to February 2026); 0.011% to 0.011% (April to May 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | 0.011% | 524 of 708 | 2 of 28 |
| Released March 2025 | 0.012% | 484 of 703 | 2 of 28 |
| The week of Aug 4, 2025 (released September 2025) | 0.0017% | 648 of 655 | 1 of 28 |
| The week of Nov 13, 2025 (released January 2026) | 0.0087% | 500 of 672 | 3 of 28 |
| The week of Feb 5, 2026 (released March 2026) | 0.011% | 484 of 677 | 3 of 28 |
| April 2026 (released June 2026) | 0.011% | 395 of 434 | 1 of 28 |
| May 2026 (released June 2026) | 0.011% | 410 of 447 | 1 of 28 |
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 59% of this job's working time as physical work. For 59% of the working time, a robot has been shown doing most of the task somewhere: 6% in a purpose-built cell, 53% in a structured facility. 1.12 of 3, the 75th highest of the 923 jobs.
- Not physical 41%
- In a purpose-built cell 6%
- In a structured facility 53%
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
51% of the working time: the conversation, in-person and hands-on parts of the tasks.
- Count cash for shipment, the hands-on work, 8%
- File slips and checks, the hands-on work, 5%
- Count the daily cash inventory, the hands-on work, 4%
- Balance the drawer, the hands-on work, 4%
- Sort cash by denomination, the hands-on work, 4%
- Take deposits, the hands-on work, 4%
What becomes checking the machine
43% 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 5% is desk work that is small if wrong, which can be handed over with lighter checks.
- Find balancing errors, costly if wrong, 7%
- Process deposits and contributions, costly if wrong, 6%
- Maintain loan records, costly if wrong, 4%
- Balance the drawer, costly if wrong, 4%
- Examine checks, costly if wrong, 2%
- Resolve account discrepancies, 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.
- Running AI inside the business systems. About 40% 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 39% of the time is hands-on work, which no model here does; robots are rated separately on this page.
- 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.
- Dealing with people outside the organization. Customers, patients, suppliers or the public: trust with them is built by a person.
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 . 28 tasks from .
| Task | ||||
|---|---|---|---|---|
| Count currency, coins, and checks received, by hand or using currency-counting machine, to prepare them for deposit or shipment to branch banks or the Federal Reserve Bank.Counting and bundling by hand and machine.Done hourly or more by most who do it. Robots: in a structured facility, Glory GLR Series teller cash recyclers (sold). Anthropic's rating. | 8% | 17 min8 min to 38 min | Hands-on 100% | Stays with people |
| Balance currency, coin, and checks in cash drawers at ends of shifts and calculate daily transactions, using computers, calculators, or adding machines.Counting the drawer and matching it to the system totals.Done daily by most who do it. Robots: in a structured facility, Glory GLR-series teller cash recyclers (sold). Anthropic's rating. | 8% | 17 min8 min to 33 min | Systems 50%, hands-on 50% | 50%Desk part half by now |
| Identify transaction mistakes when debits and credits do not balance.Working back through the entries to find the mistake.Done daily by most who do it. | 7% | 21 min8 min to 50 min | Documents 50%, systems 50% | 100%Half by now |
| Process transactions, such as term deposits, retirement savings plan contributions, automated teller transactions, night deposits, and mail deposits.Entries in the account system.Done daily by most who do it. Robots: in a structured facility, NCR SelfServ ATM with Scalable Deposit Module (SDM) / deposit automation (deployed). Anthropic's rating. | 6% | 13 min4 min to 25 min | Systems 100% | 100%Half by now |
| Receive and count daily inventories of cash, drafts, and travelers' checks.Counting the cash and instruments and recording the totals.Done daily by most who do it. Robots: in a structured facility, Glory GLR Series teller cash recyclers (GLR-100/GLR-200) (sold). Anthropic's rating. | 6% | 21 min8 min to 42 min | Hands-on 80%, systems 20% | 20%Desk part half by now |
| Receive checks and cash for deposit, verify amounts, and check accuracy of deposit slips.Counting the deposit, checking the slip, and posting it.Done hourly or more by most who do it. Robots: in a structured facility, Envelope-free deposit-automation ATMs (Peoples Exchange Bank ATM fleet) (deployed). Anthropic's rating. | 6% | 6 min3 min to 13 min | Hands-on 50%, systems 30%, in person 20% | 30%Desk part half by now |
| Sort and file deposit slips and checks.Sorting and filing paper.Done daily by most who do it. Robots: in a purpose-built cell, Panini Vision X / Vision X P (sold). Anthropic's rating. | 5% | 13 min4 min to 25 min | Hands-on 100% | Stays with people |
| Explain, promote, or sell products or services, such as travelers' checks, savings bonds, money orders, and cashier's checks, using computerized information about customers to tailor recommendations.A word with the customer at the window, with their profile on the screen.Done daily by most who do it. | 5% | 17 min4 min to 33 min | In person 70%, systems 30% | 30%Desk part half by now |
| Resolve problems or discrepancies concerning customers' accounts.Working through the account with the customer and fixing it in the system.Done daily by most who do it. | 5% | 17 min8 min to 38 min | Systems 50%, conversation 30%, in person 20% | 50%Desk part half by now |
| Process and maintain records of customer loans.Entries in the loan system.Done daily by most who do it. | 4% | 33 min13 min to 1.1 h | Systems 100% | 100%Half by now |
| Perform clerical tasks, such as typing, filing, and microfilm photography.Typing, filing and imaging.Done daily by most who do it. AI observed on this task.Robots: in a purpose-built cell, OPEX AS7200 scanning/extraction workstation with CertainScan (sold). Anthropic's rating. | 4% | 17 min4 min to 38 min | Systems 50%, hands-on 50% | 50%Desk part half by now |
| Monitor bank vaults to ensure cash balances are correct.Counting the vault against the record.Done daily by most who do it. Robots: in a purpose-built cell, Glory GLR-100 teller cash recycler (sold). Anthropic's rating. | 4% | 25 min13 min to 50 min | Hands-on 60%, systems 40% | 40%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.
- New Accounts Clerks61% now, 61% by 2028
- Cashiers19% now, 19% by 2028
- Credit Authorizers, Checkers, and Clerks86% now, 86% by 2028
- Bookkeeping, Accounting, and Auditing Clerks88% now, 90% by 2028
- Brokerage Clerks67% now, 67% by 2028
- Securities, Commodities, and Financial Services Sales Agents69% now, 73% by 2028
Where this job works
The industry groups that employ the most of it (BLS, May 2025).
- Credit Intermediation and Related Activities (5221 and 5223 only)314,650 people
- Management of Companies and Enterprises4,340 people
- Securities, Commodity Contracts, and Other Financial Investments and Related Activities3,810 people
- Nondepository Credit Intermediation2,690 people
- Investigation and Security Services1,980 people
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 Tellers is within reach of AI now?
- An estimated 48% of the working time is within reach of AI models now and 48% 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 Tellers 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: Find balancing errors (100% of the task's time now); Process deposits and contributions (100% of the task's time now); and Maintain loan records (100% of the task's time now).
- Which tasks of Tellers stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Count cash for shipment; File slips and checks; and Sort cash by denomination. Together such tasks are about 20% of the working time, and the people and body parts of the other tasks stay with people too.
- What should Tellers 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; running AI inside the business systems; and the hands-on craft. By the end of 2028 on the long-run pace at four in five, about 51% of the working time stays with people, 43% becomes checking a model's work and 5% 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 Tellers?
- In the Claude.ai conversations Anthropic sampled for May 2026 and tied to a work task, 0.011% went to this job's tasks, the 410th largest share of the 447 jobs that showed up; 1 of its 28 tasks appeared. Anthropic's labor market measure puts the share of the job's tasks seen in real Claude use at work at 2%. 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 Tellers?
- Anthropic's robot exposure rating (September 2026) finds 59% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 59% of the working time, 53% in a structured facility. Its index is 1.12 of 3. One example: Glory GLR Series teller cash recyclers for "count cash for shipment". 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 28 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.