A survey of 400 finance chiefs conducted in September by consulting firm RGP found that just 54% of respondents expect to see “significant positive” return on investment from their AI spending within 24 months ([1]). At present, only 10% of those CFOs say they are seeing such a return in their financial results ([2]). The survey, published earlier this month, covers CFOs at companies in North America, Europe and Asia Pacific with revenues above $250 million ([3]). Its authors caution that “the 24-month AI promise is now on the record, and boards will hold CFOs to it” ([4]).
The two-year timeline highlights the growing pressure on finance leaders to justify their organisations’ AI investments. In a similar vein, a July survey of 1,505 CFOs and senior finance executives by Avalara found that 92% feel moderate or significant career pressure to prove their AI projects are delivering ROI ([5]). Many CFOs have been expanding budgets for AI even as they face demands for efficiency: Deloitte’s Finance Trends 2027 survey shows 43% of CFOs ranked embedding AI and advanced tech for automation among their top three priorities, while 34% cited driving cost efficiency among their top priorities ([6]). The RGP and Avalara findings suggest that finance chiefs are bracing for a reckoning: having championed major AI spending, they now must demonstrate tangible returns on those bets, and quickly.
Research firm Gartner is advising CFOs that they will have to change how they manage workflows in order to realise value from AI-driven efficiency improvements. In guidance released on 4 November, Gartner noted that while CFOs are under mounting pressure to show returns on AI, the productivity gains delivered so far are “too marginal to materially impact financial performance or consistently lower headcount” for most organisations ([1]). A Gartner survey of 1,303 senior functional leaders earlier this year found that 23% reported reducing headcount as a result of AI adoption, but 17% actually increased headcount to add employees with strong AI and digital skills ([2]). In other words, many of the time savings from automation are being reinvested in new capabilities, or simply absorbed back into day-to-day work, rather than flowing through to the bottom line.
Grace Gstalder, a senior analyst in Gartner’s finance practice, warns that CFOs “shouldn’t assume that AI productivity gains will naturally flow through to financial performance… [because] unless organisations deliberately redirect the capacity AI creates, most of that value will simply leak away” ([3]). To convert efficiency into tangible benefits, Gartner recommends that finance leaders set clear business outcomes for any capacity freed up by AI, incorporate these redeployment goals into team objectives and metrics, and establish processes to track how productivity improvements translate into financial results ([4]) ([5]). In short, AI-driven time savings should be treated as a resource to be actively managed towards cost reduction or growth, not an automatic return in themselves.
One example of a measured AI return comes from a mid-sized software company. On 9 November, CFO.com reported that San Antonio-based Dura Software, which acquires and operates niche B2B software firms, has used an AI assistant to dramatically speed up its internal accounting and reporting workflows ([1]). According to Dura’s Chief Financial Officer, Sloan Session, the company connected Anthropic’s Claude model to its Oracle NetSuite cloud ERP via a new AI Connector service ([2]). This setup allows the AI to retrieve financial data within set access controls, follow scripted instructions and generate outputs like journal entries and financial reports for human review ([3]).
Session said the initiative has reduced processing time for routine finance tasks, from preparing bad-debt and cash-sweep journal entries to assembling monthly reports, by more than 90% ([4]). “At this point, most of our team members feel the AI… has enabled Claude to become another team member they can work with to help compile information, improve analysis, and accelerate decision-making,” the CFO noted ([5]). Importantly, Dura’s automation has been designed to assist rather than replace staff: the AI handles the grunt work of gathering data and drafting outputs, while employees still perform reviews and make final decisions. Dura’s case underscores how a focused AI deployment, led by a finance chief with a clear performance target, can yield striking efficiency gains, even if the broader financial impact is yet to be quantified.
New analysis suggests that the economic benefits of AI are so far heavily concentrated in a small minority of companies. In April, PwC’s 2026 AI Performance study found that nearly 75% of the measurable financial benefits from AI are being captured by just the top 20% of firms ([1]). These AI leaders tend to invest more and manage differently: the report found they are about three times more likely than others to have increased the number of decisions made without human intervention, and are far more aggressive in redesigning workflows to integrate AI throughout their operations ([2]). Meanwhile, the majority of companies remain stuck in experimental mode, achieving only incremental efficiencies.
There are signs that a more strategic role for finance can help close this gap. The IBM 2026 Global CFO Study, based on a survey of 1,500 finance leaders, found that organisations led by “AI-first CFOs”, executives who drive a company-wide focus on AI integration and governance, achieved revenue growth 23% higher than peers over 2022, 2024 ([3]). This indicates that clear leadership and enterprise-wide execution, rather than just technology spend, may be the key differentiator in translating AI initiatives into real business value.