Thought Leadership | Banking and Financial Services | AI and Data Engineering

Who owns the economics of bank AI intelligence?

Most banks have no single accountable executive for AI economics. That gap explains the pilot trap more than any technology issue.

Download as PDF 29th July, 2026
element
element

Beyond just having an AI strategy, every banking institution must answer one pertinent question. Who, by name, is accountable for the economics of intelligence. It’s a vacuum many boards struggle to address.

The ‘ownership vacuum’ that keeps bank AI stuck in pilot

  • AI economics today is partially owned by IT, Finance, and Risk, and operationally owned by no single accountable executive.
  • A four-tier governance structure, board to tokenomics council to BU lead to workload team, creates clear accountability across decisions.
  • A 10-decision RACI eliminates the diffusion that produces pilot purgatory and protects boards from governance merely on paper.
  • Tokenomics requires no new department, only existing roles in finance, risk, technology, and business taking on a defined responsibility.

Why the ownership vacuum keeps 95% of bank AI in pilot

In the fifth article of the Tokenomics series, Building bank AI economics maturity in 90 days, we established that the 90-day sprint is the delivery cadence that turns tokenomics from concept into live governance. In this article, we pick up the ownership question that determines whether the sprint holds beyond cycle one. The honest reality today is that most banks have no clear answer to the ownership question. AI economics is partially owned by IT, partially by Finance, partially by Risk, partially by individual business units, and operationally owned by no one. This is the structural reason that 95 percent of AI implementations remain in pilot. There is no single accountable executive whose job depends on moving them out. The fix is not a new organization chart. It is a four-tier governance structure that distributes accountability across decisions, with one named accountable executive at the top of the tier that matters most.

The four-tier bank AI governance structure

Tier 1: Board and Group Executive

The Board, through a designated committee, receives a quarterly integrated tokenomics report. The Group Executive Committee (GEC) receives it monthly and acts on it. Decisions at this tier include the AI investment envelope, the platform investment commitment, and the appointment of the accountable executive.

Tier 2: The Tokenomics Council

We referenced the Council in the previous article (above) as the forum that reviews the first live tokenomics report in the sprint. This tier defines what that Council is. The operational governance forum that does not exist in most banks and needs to. Chaired by the accountable executive, typically a direct report to the COO, CFO, or Group CIO. Membership spans Finance, Risk, Technology, Data, Compliance, FinOps, and business unit leadership. Meets monthly. Holds decision rights over the portfolio, allocation, outcome thresholds, and escalations.

Tier 3: Business unit governance

Each major business unit has a designated Business Unit AI Lead who is accountable for the tokenomics of that unit’s workloads. A senior business leader, not an engineer.

Tier 4: Workload operating teams

As introduced in our 4th article, How banks bend the AI cost curve before competitors, every meaningful workload has three named owners: business, technical, and risk. Article 4 covered what those owners do at the workload level. This tier defines where those owners sit in the institutional structure. The three-owner discipline is the single highest-leverage governance change a bank can make. More about the tiers in the PDF version of the article.

How tokenomics augments existing roles instead of adding headcount

The most common mistake banks make when reading governance frameworks like this one is to reach for the org chart and start sketching new boxes. Tokenomics does not require a new department. It requires every role that already touches AI, in Finance, Risk, Technology, Compliance, and the business lines, to take on a clearly defined tokenomics responsibility as part of what they already do. The Chief Financial Officer (CFO) and financial planning and analysis (FP&A) team extend their remit to AI consumption. Model Risk Management extends its validation remit to generative and agentic workloads. The Chief Information Officer (CIO) owns the platform investment decision and deployment gate standards. Business unit sponsors hold outcome density to account for the workloads their units run.

The leadership question that decides everything

The first article of this series asked who, by name, is the bank’s accountable executive for the economics of intelligence. This is where that question gets answered. The question is who, by name, is the bank’s accountable executive for the economics of intelligence. If the answer is not a single name and a clear remit, the bank does not yet have tokenomics. It has AI activity. The banks that will lead this decade are those in which a named, senior, empowered executive owns the economics of intelligence with the same clarity that the Treasurer owns liquidity, the Chief Risk Officer (CRO) owns risk, and the CFO owns capital.

What the full article covers

Governance frameworks look neat on paper and messy in execution. The PDF closes that gap. Inside: the full 10-decision RACI mapping every major tokenomics decision to an accountable role across all four tiers, the standing membership of the Tokenomics Council in full, and the practical observation on why AI Centers of Excellence rarely survive the transition. Plus the four functional playbooks for Finance, Risk, Technology, and business lines, showing exactly which existing roles take on which tokenomics responsibility without a single new hire.

Isn't the Council just one more committee?

Yes, and every bank calendar already carries more committees than it can absorb. The distinction the Council draws is not that it meets, but that it holds authority spanning Finance, Risk, Technology, and the business at once. That authority gap is why 95 percent of pilots do not scale.

Three functions that carry the tokenomics mandate

Finance authority

The CFO owns the AI investment envelope, workload valuation grid, and chargeback methodology. The FinOps team formally holds the AI consumption brief across all business units.

Risk authority

Model Risk Management extends validation to generative and agentic workloads. Operational risk and compliance add AI inventory completeness and outcome density to existing control frameworks.

Business authority

The Business Unit AI Lead owns the unit's investment plan, outcome density, and chargeback line. Workload business owners answer for whether each workload is earning its keep.

How to fix the AI ownership vacuum in your bank now

  • Name the accountable executive for AI economics inside the next 90 days, with a remit document on file.
  • Stand up the Tokenomics Council with the standing membership defined above and meet monthly from day one.
  • Move the AI RACI from the org chart conversation to the operating committee conversation and lock the accountable owners.
  • Require every production workload to name three owners before deployment, no exceptions.

Series close

This is the sixth article in a seven-part series on tokenomics for banks. The final article addresses the workforce and culture shift the discipline demands, where the accountability structure this article defines meets the people who must execute inside it.

The governance questions boards shouldn’t avoid

Ownership diffused between the CFO and the CIO is why most banks stall. The RACI must name one accountable executive, typically reporting to the COO, with the other as consulted.

The Center of Excellence can survive as an engineering hub, but decision rights must move to the Council. Centers rarely hold Finance and Risk authority, where tokenomics governance actually lives.

Before. The Council needs a chair with the mandate to convene it. Forming the Council first produces a forum with no owner, which is exactly the diffusion the structure is meant to fix.

Forward-looking thoughts and compelling stories

Thought Leadership

  • Banking and Financial Services

Tokenomics: The new economic discipline for banking AI

Tokenomics: The new economic discipline for banking AI Read more  

Thought Leadership

  • Banking and Financial Services

Building bank AI economics maturity in 90 days

Building bank AI economics maturity in 90 days Read more  

Thought Leadership

  • Banking and Financial Services

From cost per token to value per token in banking

From cost per token to value per token in banking Read more  

Thought Leadership

  • Banking and Financial Services

What bank CFOs must know about AI unit economics

What bank CFOs must know about AI unit economics Read more  

You define the north star, We pave the digital path

Let's connect   
elements
elements