Accounting & bookkeepingData entry and bookkeepingHigh evidence

Stanford and MIT study: accountants using AI close the books 7.5 days sooner

Stanford & MIT study (277 accountants, 79 firms) · Small and midsize firms · Generative AI accounting software (study partner)

Reported result

AI users closed month-end books 7.5 days sooner; shifted ~8.5% of time (~3.5 hrs/week) from data entry to higher-value work; reported 21% higher billable hours; 55% more weekly client support.

The problem

“Data entry and month-end close take up time we should spend on clients.”

What they did

Accountants used generative AI for foundational accounting tasks such as data entry and transaction work.

Who reported this

Independent study. Source: Journal of Accountancy (Aug 2025).

View the original source ↗

Caveats

Academic study done with an AI software firm's data; findings are associations, and billable hours were self-reported.

Results are as reported by the source above. Pathly summarizes them in our own words and labels the strength of the evidence. Your results may differ. How we label evidence

More cases like this

See all cases
Accounting & bookkeepingMedium evidence

A CPA firm saves 60 hours a month on bookkeeping with AI automation

“We're at capacity; bookkeeping and bank reconciliations keep us from tax and advisory work, and we can't keep hiring.”Steirman CPAs · BotkeeperRead the case →Accounting & bookkeepingMedium evidence

A UK accounting firm saves about 25% on bookkeeping-heavy jobs with AI receipt capture

“Getting receipts and invoices from clients is slow, and bookkeeping entry is manual.”BKB Accountants · Dext AI AssistRead the case →Accounting & bookkeepingLow evidence

Two bookkeeping firms add clients without hiring, using AI categorization

“As we add clients, data entry grows and we'd have to hire.”Decimal and Streamline Management · Dext AI AssistRead the case →