AI beats licensed accountants on speed and accuracy, but still can't close the books without supervision
AI now beats licensed CPAs on speed and accuracy—but still needs a human to sign off. Here's what that means for accounting workflows.
Why it matters
AI has crossed a capability threshold in structured accounting tasks, outperforming licensed professionals. However, the inability to autonomously complete complex multi-step closing work (APEX Benchmark) means human oversight remains mandatory—reshaping how accounting teams will be staffed and organized, not eliminating them.
The key facts
9 to knowMercor study: current AI models outperform licensed CPAs on structured accounting tasks in both speed and accuracy
18-month improvement: models lagged CPAs significantly 18 months ago; now they lead
APEX Benchmark: no model fully completes all tasks autonomously; human oversight still required for closing the books
Implication: AI augments rather than replaces; accounting profession changing shape, not disappearing
Current AI models outperform licensed CPAs on structured accounting tasks in speed and accuracy
18 months ago, AI lagged far behind on the same structured tasks
APEX Benchmark: no model fully completes all demanded tasks
AI cannot close the books autonomously without human oversight
Source: Mercor study (no public benchmark link or detailed methodology provided)
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: According to a Mercor study, current AI models now outperform licensed CPAs on structured accounting tasks in both speed and accuracy. Eighteen months ago, they still lagged far behind. On the more demanding APEX Benchmark, though, no model fully completes all the tasks. Without human oversight, AI…