The use of AI in financial departments is surging, but workers now find themselves spending a quarter of their week verifying or correcting AI outputs.
That's according to Datarails' 2026 CFO Sentiments Survey, which focused on the impact of AI on finance departments.
Datarails found that nearly all of the 270 chief financial officers (CFO) polled were using AI in some way for finance processes. While adoption is no longer a concern, the study noted that verification burden is.
Around 96% of CFOs said they spend at least 10% of their time checking or fixing AI outputs, with 8% spending more than half their working time on such efforts.
More widely, finance teams are spending 26% of their working week checking AI output to ensure it's accurate, the survey found.
Crucially, the study noted that they're right to check. Two-thirds of CFOs said their most common source of frustration with AI is when it gives confident answers based on incorrect data.
More than half (56%), meanwhile, said they've had AI provide different answers to team members despite using the same prompt and data.
Sign up today and you will receive a free copy of our Future Focus 2026 report - the leading resource for IT decision-maker insight on priorities and investment areas in AI, security and more.
"What this survey shows is that although AI adoption is now standard for CFOs, caution hasn't disappeared," said Didi Gurfinkel, CEO and co-founder of Datarails. "Finance teams have stopped asking 'will we use AI?' and started asking 'how am I going to most effectively check its work?’”
Keeping humans in the loop
Despite concerns about AI slop and inaccuracies, more than half of CFOs expect to increase their spending on AI licenses this year, with just 7% slashing budgets.
Three-quarters feel under ‘high’ or ‘very high’ pressure to integrate AI further, but just 7% said their finance department workflows are ready for a wider implementation.
Given such challenges, it comes as no surprise then that finance leaders value a human in the loop, with just 5% saying they'd hand an AI-produced financial report to the board without human review, and just 4% saying they'd do the same with monthly reports.
The survey also asked CFOs about job losses, a recurring talking point since the advent of generative AI nearly four years ago.
Yet only 3% of respondents said they were cutting full time employees due to a shift to the technology – perhaps not a surprise given checking outputs remains manual work.
"Fears of job losses have been largely allayed, but the challenge of AI output verification is critical," said Gurfinkel. "Tackling it will allow finance teams to spend less time doing intensive checking, and let them focus instead on the strategic work they were actually hired to do."
A lack of auditability remains the main reason CFOs don't trust AI tools with mission critical tasks, the report found, with 75% of those asked naming it the biggest challenge.
That was followed by accuracy and so-called hallucinations at 71% and regulatory and compliance concerns regarding AI outputs at 54%.
Because of that, Datarails noted that a new technology category was emerging, with a third of CFOs looking to implement a "finance operating system", effectively a governed data layer for AI.
At the moment, finance teams use an average of 2.5 general-purpose models, with Microsoft Copilot the clear leader at 93%, followed by ChatGPT at 65% and Claude at 64%.