Executive summary


Finance data ownership is the wrong question for chief financial officers (CFOs) and chief information officers (CIOs). The better question is which outcome Finance and information technology (IT) will deliver together. CFOs define business meaning, CIOs run stewardship and access, and the technology or AI leader builds those rules into pipelines and agents.

In the OneStream AI and Data Governance study, 85% of CIOs say IT leads data governance and 78% of CFOs say Finance does. The study also found organizations with complete Finance–IT alignment are 5.5 times more likely to report complete trust in their data. Grant Thornton’s 2026 AI Impact Survey found only 20% of organizations have a tested AI incident-response plan. Before another AI agent accesses Finance data, Finance and IT should settle four decisions: outcome, data, validation, and a joint response plan.

Who owns Finance data? Ask a chief information officer (CIO) and a chief financial officer (CFO), and you’ll hear two confident answers. In the OneStream AI and Data Governance study of 352 Finance and IT executives, commissioned by OneStream, 85% of CIOs say IT leads data governance. Meanwhile, 78% of CFOs say Finance does.

Both sides are answering the wrong question.

Everyone wants to own the data, which keeps the conversation on functions instead of results. The better question is how you’ll deliver the outcome.

Why is ownership the wrong frame?

I see this pattern often as AI changes roles and blurs lines. Well-defined functions hold tightly to what they’ve always had: “This is mine. I own it.” My response is usually the same: stop telling me what you own, and tell me how you’ll deliver the outcome.

Ask who owns the data, and the CIO will point to the data warehouse, which is now a data cloud. Finance will point to the numbers. Each answer is reasonable, and each moves the same problem forward.

We’re applying the same thinking inside our own development teams. We’re repositioning them around outcomes instead of functions. That blurs some lines and asks people to let go of ownership in favor of the result they’re accountable for.

In the Forward Finance Index, 71% of Finance leaders consider Finance and IT strategic partners. Yet only 15% report a fully integrated roadmap with shared key performance indicators (KPIs) and accountability. Only 17% of Finance executives describe IT as a strategic partner driving cross-system innovation. Another 30% call IT a reactive partner providing ad hoc support.

Gartner points to culture. In March 2026, it surveyed 223 data and analytics leaders. Cultural resistance outweighed funding constraints as the primary reason governance initiatives fail (60% versus 40%). Gartner recommends treating data governance as a business enabler and shared responsibility, not an IT-only function.

What should you ask instead?

Start with the outcome you’re trying to drive, such as a forecast the board will rely on or a close the auditors can follow. Then the questions get practical:

  • What data do we need for this outcome?
  • How do we validate that the numbers are right?
  • Can we stand behind the insight in the report?

Finance brings the business meaning: which definition applies and when an error matters. IT brings the systems knowledge: where the number lives, how it moved, and who can change it. You need both to answer those questions, and neither covers the other’s job.

With AI, the limit isn’t what the model can do. It’s the context you give it, and that context comes from the outcome you’re trying to drive.

Why does this matter for agents?

People have traditionally sat between the data and the decision, asking questions, applying business context, and catching errors. As agents begin recommending or initiating actions, that human buffer can become thinner. An AI agent may pull information from financial systems, planning tools, and enterprise data platforms, then recommend or initiate an action. Without defined sources, permissions, business definitions, and review requirements, weak data can reach forecasts, close processes, and business decisions.

Many organizations are already confronting this risk. In Grant Thornton’s 2026 AI Impact Survey of 950 US business leaders, nearly three in four organizations give agentic AI access to data and processes. Only 20% have a tested AI incident-response plan. And 78% of executives lack strong confidence they could pass an independent AI governance audit within 90 days.

The issue is not that agents inevitably operate without controls. The issue is that controls, accountability, and response plans have not always kept pace with deployment.

Who contributes what?

Shared outcomes still need clear roles. In the Office of the CFO, the work divides by contribution, not by ownership.

  • The CFO defines the business meaning behind critical data, including KPIs, hierarchies, and materiality thresholds, and decides what qualifies as decision-ready.
  • The CIO runs stewardship, lineage, cataloging, and access, so every critical data set has a documented path from source to report.
  • The technology or AI leader, often the chief technology officer (CTO), builds those rules into pipelines and agents. That includes limiting agents to governed sources and logging what they do.

Each leader has an explicit, complementary role in the same outcome. That is different from everyone being vaguely responsible.

What do aligned teams do?

The Forward Finance Index compared the market average with Forward Finance Leaders. Leaders are the 6% of organizations that rank in the top two quartiles across all five dimensions. Three behaviors distinguish them:

  • Shared strategy: 81% of Leaders say Finance and IT co-define data and AI strategy, compared with a 66% market average.
  • Shared governance: 78% of Leaders share formal governance for data, AI, and digital platforms, compared with 66% overall.
  • IT as innovation partner: 85% of Leaders say IT helps Finance translate business needs into scalable, secure solutions, compared with a 71% market average.

Alignment also pays off in trust. In the OneStream AI and Data Governance study, organizations with complete Finance–IT alignment are 5.5 times more likely to report complete trust in their data.

What should you settle first?

Four decisions come before another AI agent gains access to Finance data or workflows:

  1. Name the outcome. Agree on the decision or report the data supports, and who is accountable for that result.
  2. Agree on the data it needs. Publish the governed sources agents may use, along with the definitions and permissions that apply.
  3. Define how you’ll validate it. Set the checks on the numbers and the human-approval thresholds: which recommendations an agent can surface, which actions it can start, and which decisions need review.
  4. Create a joint response plan. Finance and IT agree on who responds when an agent produces an incorrect answer, takes an inappropriate action, or uses the wrong source.

That final step matters most, because only 20% of organizations report having tested an AI incident-response plan, according to Grant Thornton.

The cost of getting this wrong is already measurable. In the same Harris Poll survey, 72% of executives say bad data cost their organization $500,000 or more, and 37% report losses above $1 million.

So skip the ownership debate. If your CFO and CIO each answer “I own it,” ask a different question: what outcome are we trying to deliver together?

Explore the Forward Finance Index

Jeremy Guttman is Executive Vice President of Product & Engineering at OneStream Software. With over 20 years of experience in corporate performance management (CPM) and data analytics Jeremy is a collaborative leader with proven success managing teams and building application platforms focused on the intersection of planning & data analysis. Prior to OneStream, Jeremy served in leadership roles in Product Management, Services, & Support at Tidemark prior to the acquisition by Longview Solutions. Jeremy's experience also includes Anaplan, SAP, and OutlookSoft where he held a variety of roles focused on helping Fortune 500 customers build & implement scalable business solutions.

Demo Sign Up