Will AI Replace Bank Tellers? What the Job Data Actually Shows

Largely yes, and the process started well before generative AI existed. The Bureau of Labor Statistics projects employment of tellers to fall 13 percent from 2025 to 2035, a loss of about 44,700 positions from a base of 339,200. That is one of the steepest declines in the whole occupational handbook, and it is being driven by mobile banking and self-service far more than by anything a language model does.

The useful question is not whether the role shrinks. It is what happens to the people in it, and which parts of branch work are actually growing while the transaction counter disappears.

Key points

• The decline is real and already underway. A projected 13 percent drop to 2035, roughly 44,700 fewer jobs.
• Automation here is old technology, not new AI. Cash machines, mobile deposit and self-service kiosks removed most routine transactions years ago.
• Median pay is $43,030, with a high school diploma as typical entry-level education.
• Branches are not closing as fast as teller roles are. Banks are converting counters into advisory desks rather than shutting locations outright.
• The exit route is sideways, not out. Personal banker, relationship banker, fraud operations and small business support all draw from the same pool and pay more.

What tellers actually do, and which parts survive

The job is usually described as processing deposits and withdrawals. In practice a teller handles four distinct kinds of work, and they have very different futures.

Routine transactions. Deposits, withdrawals, transfers, cheque cashing. This is the part that machines took. A customer who once queued to move money now does it on a phone in fifteen seconds, and the volume that remains at the counter is a fraction of what it was.

Cash handling and balancing. Counting, verifying, managing a drawer, reconciling at end of day. Partly automated by recyclers and counting machines, still requiring a person on site wherever physical cash exists.

Identity, judgement and exceptions. A customer whose card is blocked, a cheque that looks wrong, a large withdrawal that does not fit a pattern, an elderly customer being coached by someone on the phone. This is where a teller catches things no system flags, and it is the part of the job with the highest value per minute.

Conversation and referral. Noticing that someone mentioning a house purchase should talk to a mortgage adviser, or that a small business owner is using the wrong account type. Banks describe this as the reason branches still exist, and it is the only part of the job that generates revenue rather than processing it.

The split matters because national projections average across all four. A teller whose branch still runs high transaction volume is in a very different position from one whose day is already mostly exceptions and referrals, even though the statistics treat them as the same occupation.

Where AI is genuinely changing the work

The technology arriving in branches now is not replacing the counter, which was already largely gone. It is changing the surrounding work.

Fraud and anomaly detection runs continuously across transactions and flags patterns a person would never see. Identity verification increasingly uses document scanning and biometric checks rather than visual inspection. Customer service chat handles the routine questions that used to arrive by phone. Internal assistants help staff find the right policy or product answer without leaving the customer waiting.

Notice what these have in common. They reduce the volume of simple work reaching a human and raise the proportion of interactions that are complicated, emotional or unusual. That is a harder job, not an easier one, and it is why the roles replacing teller positions tend to pay better.

There is a second-order effect worth naming. When a system handles everything straightforward, the human sees only hard cases all day, with no easy ones in between to reset on. Contact centres discovered this before banks did: removing routine work raises the average difficulty of every remaining interaction, and staffing has to account for that rather than simply cutting headcount in proportion to volume.

Why the decline is not the same as elimination

Three things keep people in branches even as the teller count falls.

Cash has not disappeared. Handling, securing and reconciling physical currency requires staff on site. Cash use declines slowly and unevenly, and it remains important for older customers, small businesses and lower-income households.

Regulation requires human decisions. Anti-money-laundering rules, suspicious activity reporting and know-your-customer checks all involve judgement that a bank must be able to defend. Software surfaces the alert; a person decides and signs.

Trust is the product. For a mortgage, a bereavement, a suspected scam or a business loan, customers want a person. Banks that stripped branches back too aggressively have reversed course in several markets, and the roles they rehired for were advisory rather than transactional.

A fourth factor gets less attention than it deserves. Fraud against retail customers has grown more sophisticated, and a good deal of it depends on the victim being isolated from anyone who might ask an obvious question. Branch staff intercept a meaningful share of authorised push payment scams simply by noticing that a customer is being talked through a large transfer by someone on the phone. No detection system currently replicates that, because the transaction itself looks entirely legitimate; what gives it away is the customer’s demeanour.

The pattern to hold onto is that the counter is going away while the branch adapts. If you are currently a teller, the risk is not that banking stops needing people. It is that it stops needing the specific task you were hired for, on a timeline measured in a few years.

What to know before deciding

For comparison, the wider picture is not this bleak. The Bureau of Labor Statistics projects total employment to grow from 170.3 million to 176.2 million between 2025 and 2035, an increase of about 5.9 million. Teller work is declining against a growing labour market, which means the people leaving it are moving into other jobs rather than into unemployment.

The shape of the decline matters as much as its size. Losing 44,700 positions over ten years across a country works out to a few thousand a year spread over thousands of branches, which in practice means a departing teller is not replaced and their hours are absorbed by colleagues. Very few people will be told their job has been automated. Many more will find that the vacancy they hoped to move into was quietly removed from the org chart, which is a slower and less visible thing to plan around.

It is also worth being precise about the AI claim. The Bureau publishes AI exposure categories for 831 occupations and states directly that exposure “does not imply job loss, productivity gains, automation probability, or wage effects.” Teller decline is well documented and predates the current wave of AI. Attributing it entirely to chatbots gets the cause wrong and, more importantly, points people toward the wrong response.

Roles that hire tellers, ranked by how far the skills carry

• Personal or relationship banker. The most direct step. Same customers, same products, more advisory work and better pay.
• Fraud and financial crime operations. Tellers who are good at spotting things that do not add up transfer well, and this function is growing.
• Small business banking support. Requires learning products rather than a new field, and demand holds up.
• Contact centre and digital support. Same customer knowledge applied through different channels.
• Branch operations and compliance. Suits people who like the procedural side of the role, and it is one of the few branch functions with stable or rising headcount.
• Wealth and mortgage support roles. Longer training and licensing, but the pay gap over a teller position is substantial and the demand is not falling.

Two things are true of every item on that list. Each pays more than a teller position, and each requires knowledge a teller already has about products, customers and how a branch actually runs. That existing knowledge is the reason internal moves work better than starting over in an unrelated field, and it is why employer-funded training is the route worth taking seriously.

Decision framework

If you are in this job now, or considering it, work through these five.

1. How many years do you need from this role? A decade is a problem. Two or three years while you build toward something else is manageable, since the decline is gradual rather than sudden.
2. Is your branch converting or closing? A branch being redesigned around advisory desks will retrain staff. One scheduled to close will not.
3. Which of the four task types do you spend most time on? If it is routine transactions, you are exposed. If it is exceptions and referrals, you are already doing the work that survives. Time yourself over a week rather than estimating, because most people are surprised by the split.
4. Do you have a licence or certification path available? Many banks fund the qualifications needed to move into advisory roles, and taking that offer is the single highest-return move available to a teller.
5. Can you learn the systems rather than just operate them? Understanding how fraud flags are generated, and where they are wrong, turns you into the person who reviews alerts rather than the one replaced by them.

Getting comfortable with how these systems reason, and where their output needs a human check, is a transferable skill rather than a bank-specific one. If you want a structured route in, explore Coursiv AI lessons and check current plan details on the official site.

Your next step

If you are a teller, audit one week of your own work and write down how much time goes to routine transactions versus exceptions, referrals and judgement calls. That ratio is your personal exposure, and it is far more informative than any national projection.

Then find out what your employer will fund. Banks are actively retraining branch staff into advisory roles because they need those people, and the qualification they pay for is worth considerably more than the same course bought privately.

FAQ