Finance professionals spend up to half their workweek verifying AI outputs, survey finds

A new survey from accounting ERP provider Sage reveals that the time-saving promise of artificial intelligence is being eaten away by the hours finance professionals must spend double-checking the technology's work. According to Accounting Today, which reported on the survey, the findings point to a growing productivity paradox where the overhead of auditing AI offsets the efficiency gains it offers.

"AI's time-saving benefits are being significantly eroded by the time required to verify AI outputs."

That erosion is stark in the numbers: 48% of finance professionals surveyed spend 15 or more hours per week on verification activities, and 19% devote over 30 hours weekly to that task, Accounting Today reported. Additionally, 26% of respondents spend more than a quarter of their time simply figuring out how AI models reached their conclusions so they can explain them to others.

Separate analysis from the technology site Nexairi puts the toll in percentage terms, noting that about 37% of AI's time savings goes back into checking its work. The same piece suggested that finance teams can preserve gains by "tier[ing] review by consequence so verification does not cancel the gain," a structured approach that prioritizes oversight based on the impact of a given output.

"The findings highlight a growing productivity paradox where the overhead of auditing AI work offsets the efficiency gains AI promises."

The Sage data underscores that simply deploying AI tools is not enough; organizations must also build systems to manage the verification burden. Without such steps, the hours saved by automation risk being swallowed by the very act of ensuring the automation is correct.