After two years of pilots, some CFOs are looking beyond hours saved to AI’s potential impact on cash flow.
For most of the past two years, the case for AI in finance centered on efficiency: closing the books faster, cutting manual reconciliation and moving analysts off spreadsheets. Adoption followed. In a June 2025 Gartner survey of 183 CFOs, 84% said they had deployed or planned to deploy AI.
Results have not kept pace. In the same survey, only 7% of CFOs reported a high impact on their function. The gap between adoption and outcome raises questions about how finance leaders evaluate the technology, including its impact on cash flow.
Why hours saved fall short
Time savings are hard to bank. When an analyst spends four fewer hours on reconciliation, the savings do not appear on the income statement unless the company changes headcount, scope or speed. Cash does appear. Faster collections, lower days sales outstanding (DSO) and fewer disputed invoices move working capital directly, and a CFO can take those numbers to a board.
“We’re way past the stage of ‘let’s try some AI and pilot a thing,'” said George Kurdin, co-founder and CEO of Monk, an AI accounts-receivable company. “Now it’s about hard ROI.”
The view from Net 30
That shift was on display on October 21 at Net 30, an invite-only summit on AI and finance hosted by Monk at The Times Center in New York. The event was advertised as a gathering of about 150 participants.
The evening ran under one tagline: cash is the new growth metric. Revenue growth tells only part of a company’s financial story. Today, investors and boards look harder at how quickly revenue turns into cash, and Net 30 put that question at the center of the AI conversation.
Speakers approached the same question from different angles. According to Monk’s account of the program, Damian Lewandowski of McKinsey explained why many agentic AI pilots stall, noting that while experimentation is widespread, few companies are banking the value. Ryan Farmer of Vestis outlined the metrics he expects finance teams to be measured on by 2027, with cash and DSO near the top. Itay Goldstein of the Wharton School addressed the risk of letting autonomous agents act on financial data without guardrails.
A closing panel of operators, Jason Pratt of Revvity, Adam Klein of Metronome, Rahul Malhotra of Integreon and Keith Weissman of Teladoc Health, focused on what AI has changed inside their organizations.
Attendee feedback shared by Monk pointed to the same theme. Several guests said the distinction between hours saved and cash in the bank was the idea they would bring back to their teams. One said a chart comparing compute cost with realized benefit changed how he would present AI projects internally.
Measuring returns in accounts receivable
Accounts receivable is one of the clearest places to measure AI’s return. It sits close to the bank account, its performance is already tracked through DSO and collection rates, and much of the work is repetitive: sending invoices, following up on late payments, matching remittances and resolving disputes.
Monk says its AI agents help automate collections work. The company’s position is that AI in finance should be evaluated like any other investment, by its effect on cash flow. Faster payment and fewer write-offs are outcomes a finance team can verify directly, without a separate model to estimate time saved.
The next test
The guardrails question raised at Net 30 is one consideration as AI agents take on more of the finance workflow. CFOs will need confidence that these systems act within policy, keep clean audit trails and escalate exceptions to people. Those controls should be evaluated alongside any demonstrated cash-flow benefits.
For finance leaders, the next round of AI budget discussions is likely to turn less on what the technology can do and more on what it has returned. If cash is the new growth metric, that is where AI will be judged.
Following the response in New York, Monk plans to take Net 30 to more cities next year, with details to be announced.
