In nearly every AI governance conversation I have with leadership teams, the CFO is either absent or an afterthought. The CIO owns the technology. The Chief Data Officer owns the data. Finance gets looped in when it is time to approve a budget.
That structure feels logical. It is also, according to the data, one of the most expensive assumptions organizations are making right now.
Our Q1 2026 report, “Economic Maturity for Artificial Intelligence,” puts a number on the cost of that assumption. When CFOs own AI accountability, 76% of organizations achieve high AI value. That drops to 53% under CIOs or CTOs, and 32% under functional executives. Only 2% of companies have actually made that decision.
Why finance outperforms technology in AI governance
The CFO’s advantage is not about understanding algorithms. Most CFOs would readily admit they are not AI technologists, and that is precisely the point.
Finance brings three capabilities to AI governance that technology leaders, despite their expertise, are rarely positioned to deliver:
Measurement discipline. Finance teams exist to challenge numbers. They are trained to ask whether a claim is real, repeatable, and attributed correctly. When an AI project shows promising results, a finance lens asks: how much of that outcome was actually a result of the model? What would have happened without it? Those questions are the foundation of accountability.
Certification, not just approval. The research is specific on this point. Executives interviewed for this study described finance partnering with technology leaders specifically to certify AI value, not just approve AI budgets. This is an important distinction. Budget approval is a gate. Value certification is an ongoing process. Finance brings the latter in a way that most AI governance structures never require.
Cross-functional credibility. When the CFO signs off on an AI initiative’s outcomes, the organization believes it. Sales, operations, marketing, supply chain: each of those functions has reason to trust a number that has been validated by finance. That credibility accelerates adoption and sustains investment in ways that technology-led governance rarely achieves on its own.
The measurement gap is where AI value goes to die
Here is a pattern I have seen repeatedly across organizations at very different stages of AI maturity: they invest in AI, they run pilots, they see encouraging results, and then the value disappears somewhere between the pilot and the profit and loss statement.
The culprit is almost always measurement. Not measurement of whether the model worked technically. Measurement of whether the business outcome changed, by how much, and because of what.
Finance is the natural owner of that infrastructure. Attribution modeling, portfolio reporting, assumption validation, outcome verification: these are the core skills of the finance function, applied to a new class of investment.
What 2% tells us about where most organizations land
Only 2% of organizations have assigned CFO ownership of AI accountability. That number deserves a moment of consideration.
It is not that finance leaders are unavailable or uninterested. In my conversations with CFOs across industries, many are actively trying to get closer to AI strategy and find their organizations are structuring them out of it. The assumption that AI belongs to the CTO, with finance brought in at the end to approve a budget, is deeply embedded in how companies are organized.
The consequence is predictable. AI projects get approved, sometimes generously. But value certification never happens. There is no mechanism to ask whether the investment actually returned what it promised. By the time the next budget cycle arrives, the organization is approving new AI investments without a clear accounting of what the previous ones delivered.
At its core, this is a governance problem with a financial solution.
DBS Bank shows what CFO accountability produces
One organization has already figured this out. According to a March 2026 Fortune report, unit CFOs at DBS Bank in Singapore are responsible for vetting AI value numbers before they roll up to the enterprise level.¹
The results speak for themselves. DBS generated SGD 370 million in economic value from AI in 2023. That more than doubled to SGD 750 million in 2024. The bank’s CEO has since indicated they are on track to reach between 1.1 and 1.2 billion Singapore dollars in 2025.²
The compounding growth reflects what happens when the finance function holds real accountability for AI outcomes, changing what gets counted, what gets challenged, and ultimately what gets delivered.
AI governance follows the same logic as capital investment
Think about how a major capital expenditure works in a well-run organization. Engineering specifies what is needed. Technology evaluates the options. Finance validates the business case, monitors the deployment, and signs off on whether the project delivered its projected returns.
Nobody argues that the CFO should have run the engineering analysis or selected the vendor. The value finance adds is fiduciary, and the same logic applies to AI.
The CTO should absolutely lead on architecture and build. The CDO should own data strategy and use case development. The CFO should own accountability for whether the investment returns value. These are complementary roles, not competing ones.
The organizations achieving high AI value appear to understand this. The other 98% are still figuring it out.
The practical question
If you are in a leadership position and your organization has not made this decision, the question worth asking is not whether the CFO is technically qualified to oversee AI. The question is whether your current governance structure has anyone whose job is to certify that AI investments are delivering real business value.
If the honest answer is no, then the data offers a clear direction.
The finance function has been measuring and validating business outcomes for a long time. AI is a new type of investment, but the discipline required to hold it accountable is not new at all.
Laks Srinivasan is co-founder and CEO of the Return on AI Institute (RoAI Institute), which helps leadership teams build the AI fluency required to make confident, consequential decisions.
The CFO accountability finding is one of several patterns from our Q1 2026 survey that point toward governance and measurement discipline as the differentiating factors in AI value creation. The full findings are in our report, “Economic Maturity for Artificial Intelligence,” co-authored with Thomas H. Davenport. We analyzed these patterns across more than 1,000 C-suite executives in 11 countries and 32 industries. If your leadership team is working to close the gap between AI investment and confirmed return, the report is worth your time. Download it here.
¹ Loten, A. (2026, March 27). Why CFOs—not chief AI officers—are the secret to getting real value from AI.Fortune. https://fortune.com/2026/03/27/why-cfo-not-chief-ai-officer-secret-getting-real-value-ai/
² DBS Bank. (2024). DBS Annual Report 2024. https://www.dbs.com/annualreports/2024/; McKinsey & Company. (2025, September 23). DBS CEO Tan Su Shan on building a gen AI-enabled bank with a heart.https://www.mckinsey.com/featured-insights/future-of-asia/dbs-ceo-tan-su-shan-on-building-a-gen-ai-enabled-bank-with-a-heart
