AVGPAY AI & WORK RESEARCH

How could AI change the work of bookkeeper?

The evidence points to high modeled task exposure (56/100) for this occupation. This does not mean a 56% chance of losing a job. Compare what the research can actually measure with typical pay, published demand and the work people do.

AI task exposure
56 / 100
High · 2023 study assumptions
National median pay
$50,670
OEWS published wages
Employment change
-5.6%
BLS 2025–2035 projection
Average annual openings
144,100
Includes replacement openings

How this occupation compares

U.S. weighted average: 33Office & admin field: 57

Bookkeeper has a score of 56, close to the office & admin field average of 57. Its score is higher than occupations employing approximately 84% of workers in the covered wage dataset. These are job-weighted comparisons, not a survey of worker AI usage.

Tasks potentially accelerated by a chatbot alone
37%
With supporting software and tools
74%

The model assumes some tasks could be completed at least twice as quickly with comparable quality. It does not say employers have automated those tasks or eliminated equivalent numbers of jobs. The two scenarios are modeled capabilities from the study's historical baseline, not measurements taken in 2026.

What work could change—and what still needs investigation?

Compute, classify, and record numerical data to keep financial records complete. Perform any combination of routine calculating, posting, and verifying duties to obtain primary financial data for use in maintaining accounting records. May also check the accuracy of figures, calculations, and postings pertaining to business transactions recorded by other workers. The activities below are documented occupation tasks, not a list of tasks that AvgPay has individually proven AI can automate.

  1. Check figures, postings, and documents for correct entry, mathematical accuracy, and proper codes.Which information-handling steps could AI assist with, and which need human verification?
  2. Operate computers programmed with accounting software to record, store, and analyze information.Which information-handling steps could AI assist with, and which need human verification?
  3. Comply with federal, state, and company policies, procedures, and regulations.What judgment, interaction or domain expertise is necessary to complete this task well?
  4. Debit, credit, and total accounts on computer spreadsheets and databases, using specialized accounting software.Which information-handling steps could AI assist with, and which need human verification?
  5. Classify, record, and summarize numerical and financial data to compile and keep financial records, using journals and ledgers or computers.Which information-handling steps could AI assist with, and which need human verification?
  6. Code documents according to company procedures.Which information-handling steps could AI assist with, and which need human verification?
  7. Access computerized financial information to answer general questions as well as those related to specific accounts.Which information-handling steps could AI assist with, and which need human verification?
  8. Operate 10-key calculators, typewriters, and copy machines to perform calculations and produce documents.Which information-handling steps could AI assist with, and which need human verification?

Three practical questions help turn a general exposure score into a meaningful discussion: which tasks involve routine information processing; which need confidential data, legal responsibility or face-to-face judgment; and which parts of the workflow have actually adopted AI in your industry? The study cannot answer those for your employer.

Skills and preparation that shape the decision

These are important occupational skills reported in the skills inventory. They are not evidence of immunity to automation or recommendations to abandon existing expertise.

  • Mathematics
  • Active Listening
  • Critical Thinking
  • Reading Comprehension
  • Speaking

Typical entry education: Some college, no degree. Related experience: None. On-the-job training: Moderate-term on-the-job training. A transition may require credentials, licenses and specialized experience even when two jobs appear similar.

BLS projects a decrease of 5.6% in occupation-level employment between 2025 and 2035, from approximately 1,532,400 to 1,446,800 positions. Its annual openings measure includes people leaving an occupation, which explains why openings can exist even with negative net growth. The projection is not a causal estimate of AI job displacement.

Compare other careers before making a move

These are neighboring field occupations with at least 15 points lower modeled exposure and broadly comparable median pay. They are not verified career transitions; look carefully at required training and the actual tasks.

OccupationAI exposureMedian payResearch
Mail Carrier29/100$60,550Explore
Mail Sorter11/100$58,470Explore
Postal Clerk30/100$62,130Explore
Weigher28/100$46,380Explore
Meter Reader40/100$48,150Explore

So, is this job likely to disappear?

That cannot be determined from the current exposure score. Exposure measures theoretical task acceleration, not probability of replacement, employment effects, business investment or the pace at which a profession adapts. The best decision is to compare this occupation's current tasks with the latest BLS growth outlook, then investigate actual AI adoption where you work.

This page includes a source-backed task inventory, occupational outlook and direct exposure study match.

Full salary, forecast scenarios and career paths → · Search another occupation → · Workforce research report →

Methods: AvgPay exposure methodology. Original study: Eloundou, T., Manning, S., Mishkin, P. and Rock, D. GPTs are GPTs: Labor market impact potential of LLMs. Science (2024); working paper arXiv:2303.10130 (2023).. Employment and salary sources are documented in each career profile.