Are CFOs ready to govern AI investment? IBM finds a wide execution gap
A survey of 1,500 finance leaders across 33 geographies finds broader authority over AI strategy, but only 6% describe finance as transformation-ready. The results map perceptions and associations—not audited returns or proof that AI caused better performance.
The short version
- What happened
- IBM and Oxford Economics surveyed 1,500 CFOs or equivalent finance leaders across 33 geographies and 26 industries from February to April 2026.
- Who it affects
- Finance staff may spend less time assembling routine forecasts but face greater responsibility for checking data, assumptions and automated recommendations.
- What remains unproven
- The findings are based on a survey of senior finance leaders and a composite using self-assessed performance rather than independently audited business outcomes.
Keep in mind
The findings are based on a survey of senior finance leaders and a composite using self-assessed performance rather than independently audited business outcomes.
The public release does not report the response rate, sampling frame, weighting, subgroup denominator, confidence intervals or full question wording.
IBM sells AI and consulting services, so its commercial position should be considered alongside the disclosed method and results.
Source analysis · 2 linked sources. Primary analysed source date: 30 September 2026.
Read the full report · 5 min The evidence, methods, implications and limitations
A broad survey finds authority expanding faster than operating maturity
IBM's Institute for Business Value, working with Oxford Economics, surveyed 1,500 chief financial officers and equivalent senior finance leaders between February and April 2026. Respondents came from 33 geographies and 26 industries. In the results released on 30 September, 62% said their role had expanded into enterprise technology or AI strategy leadership, 56% reported greater authority over portfolio management and capital reallocation, and 54% said they had taken on more responsibility for business-model or growth-strategy design.
The maturity picture is much narrower. Only 6% described finance as transformation-ready, meaning that AI was consistently embedded in workflows and decision-making at scale. Forty-eight per cent placed finance in a developing stage where AI skills remain concentrated in selected roles, teams or use cases. Another 42% reported a high stage of readiness with AI-fluent teams capable of scaling. Those categories are useful signals of confidence, but the public release does not show the exact question wording or an independently tested maturity score for each organisation.[1][2]
Capital controls remain mostly human and only partly connected to value
Almost half of respondents—48%—said their organisations frequently update capital allocation for AI and growth investments using real-time, data-driven insights. The same share said finance actively tracks AI-driven value creation and reallocates capital accordingly. Yet only 8% said finance leads enterprise-wide AI value goals and has automated investment triggers tied to those results, while just 6% allow AI to recommend or execute reallocations within defined guardrails. The survey therefore does not describe finance handing control of budgets to autonomous systems; it describes selective automation inside largely human approval structures.
For boards and investors, the practical question is what sits behind 'AI-driven value'. A defensible investment case separates revenue, avoided cost, cycle-time improvement and risk reduction; records the baseline and total cost; and assigns an accountable owner. It also includes failed pilots, integration costs, data work, model monitoring and staff time. Without that discipline, a faster approval process can simply move money more quickly, while a dashboard labelled real-time can still optimise the wrong business measure.[1][2]
The higher-performance subgroup does not establish cause and effect
IBM identifies an 'AI-first CFO' group whose organisations report advanced capabilities across strategy, governance, integrated intelligence, capital allocation and long-term planning. The release says organisations led by this group had revenue-growth rates 23% higher relative to peers from 2022 to 2024, approved new AI funding 15% faster and were 18% more likely to report effective strategy execution. Those comparisons may help form hypotheses, but they do not prove that a CFO's AI posture produced the differences.
The methodology says a composite performance measure used respondents' assessments of growth, efficiency and productivity, agility and adaptability, and risk preparedness relative to competitors. That makes the analysis vulnerable to common-source and self-assessment bias: confident leaders may rate both their AI capabilities and their performance highly. Company size, sector, prior digital investment, profitability and management quality could also influence both. The public summary does not provide subgroup size, response rate, weighting, confidence intervals or a causal identification strategy.[1][2]
What finance teams can do now—and what would change our assessment
A useful response is governance that matches decision risk. Low-stakes drafting or reconciliation tools can be tested with routine review, while forecasting, credit, pricing or capital-allocation systems need traceable data, documented assumptions, access controls, scenario tests and a named human decision-maker. Finance should keep a register of AI projects, their intended outcome, cost, owner, model and data dependencies, approval boundaries and evidence after deployment. Staff need room to challenge recommendations rather than being measured on whether they accept automation.
Our assessment would strengthen with the full questionnaire, sampling and weighting details, the number of CFOs in the AI-first group, objective company-level outcome data and a longitudinal test showing whether governance changes precede improvements. It would weaken if the subgroup were small, categories were chosen after examining outcomes, or independent data failed to reproduce the reported advantage. The survey is strong evidence that many finance leaders perceive their remit expanding; it is not evidence that most finance functions have operationalised AI successfully or that AI caused superior returns.[1][2]