Late last month, Apollo Global Management sounded an alarm that outside of a handful of tech giants, artificial intelligence has so far delivered “zero financial payoff to corporate America” ([1]). As surging cloud and “token” costs have ramped up spending, finance chiefs have become more hard-nosed about funding AI experiments. Evidence is growing that CFOs are rethinking big bets on generative AI and refocusing on measurable outcomes.
At the Digital Finance Forum in New York on 17 November, Eric Glyman, CEO of expense management platform Ramp, said his company’s data shows a major shift in how firms allocate AI budgets ([2]). He noted that many CFOs have moved on from the “blank-check” approach that marked early generative AI pilots and are instead channelling investment into use cases with a clear business case and trackable benefits. This marks a pivot from the hype-driven projects of a year ago towards tighter cost discipline and proof of value.
UK companies have rapidly established Chief AI Officers (CAIOs) to help drive value from AI. According to a new IBM Institute for Business Value study released on 17 November, 78% of large UK organisations now have an executive in the CAIO role, up from just 26% in 2025 ([1]). These specialists are tasked with identifying where AI can improve efficiency, revenue or margins. But IBM found a critical disconnect: only 16% of surveyed CAIOs have full control over AI investment decisions ([2]). Just 32% say their organisations can quickly reallocate funding when AI performance data points to better results ([3]).
This gap between insight and authority means that many AI opportunities risk stalling before they affect the bottom line. Even though 34% of UK AI chiefs say the technology is already reshaping their business models ([4]), most lack direct budget power to act on those insights. The study suggests that companies could be leaving value on the table if the people who understand how to capture ROI from AI are not empowered to adjust spending. CFOs may need to work closely with their AI leaders, or take on the role themselves, to ensure that high-impact AI initiatives receive the necessary funding and oversight.
New research indicates that realising AI’s financial returns requires more than automating tasks or cutting costs, it demands active management to convert those benefits into results. In a Gartner survey of 1,303 senior functional leaders conducted earlier this year, more than a quarter of finance departments reported reducing headcount due to AI, while only 6% saw a staff increase ([1]). Finance teams also reported smaller productivity improvements from AI (19% on average) than other functions such as IT or marketing (22%) ([2]). This reinforces the view that efficiency gains, on their own, are not yet translating into higher profits for most firms. A separate Gartner survey of 275 customer-service leaders found that a mere 4% achieved the ROI they expected from AI projects ([3]), underscoring how uncommon a full return on AI investment remains in practice.
Gartner analysts advise CFOs to treat AI as a catalyst for transforming work, rather than a simple headcount reduction tool ([4]). Crucially, they warn that time or cost savings from AI will not hit the profit-and-loss statement unless those gains are reinvested in activities that drive performance ([5]). In research presented on 16 November, Gartner noted that finance employees freed up about five hours per week through AI automation, yet only 1.7 of those hours were reallocated to higher-value tasks ([6]). To avoid having the rest of that capacity "simply leak away" ([7]), CFOs are being urged to set clear targets for using AI-driven efficiency gains, for example, redeploying staff to improve forecasting or decision support, and to track the impact on costs or revenue. The message: without deliberate action, even genuine productivity improvements from AI risk fading into unmeasured work instead of visible ROI. A recent global survey by KPMG similarly found that only 12% of organisations consistently assess AI’s value against its cost, though nearly half of the few reporting established returns do so ([8]). The drive to realise AI ROI is pushing finance leaders to implement more rigorous cost governance, benefit tracking and strategic alignment of AI initiatives with business goals.