Interim Chief AI Officer — Value Agenda

Paying for AI by what it changes

Artificial intelligence budgets are easy to grow and hard to justify. Use cases multiply because each is cheap to start, and the cost appears later, in compute, in data work, in model maintenance and in the operational risk of a model behaving in ways nobody expected. The Interim Chief AI Officer, who reports to the Technology Committee Chair, is appointed to tie AI and product funding to two things over 24 months: measurable economics and technical consequence. A use case earns funding when the officer can show what it does to cost, revenue or service, and what it commits the architecture to.

The seat carries executive authority. The officer decides how large an AI or product investment local leaders may approve alone, and specifies the proof required before AI spending, compute and engineering capacity, or model controls change. They judge which existing commitments earn their place and which should pause or close. They block proposals that offload cost or risk onto teams that never accepted it, decline value cases that skip adoption, upstream dependencies or the downside, confirm who leads AI work and how problems are raised, and settle disagreements with operating leaders while budgets and staff are still uncommitted. They also decide how often the board hears about model performance, exceptions and decisions still open. Model risk sits squarely inside all of this, because a model that routes freight or prices a contract carries consequences a pilot dashboard never will.

Two years from portfolio to funding discipline

The first ninety days should establish the AI use-case portfolio as it really stands, agree the criteria for funding, and put model risk controls in place, with a single accountable lead for each use case. Around month nine, AI and product funding should be tied to economic and technical evidence, and the weaker use cases should have been paused or closed, however popular they were. By month 24, the business should be running a funding cycle that weighs economics and technical consequence as routine, with AI governance robust enough that its results do not depend on the interim officer being in the room.

An officer who has scaled before

You have carried a scale-up that was felt in the boardroom and on the operating floor alike, in a business no smaller than $335 million in revenue and 1,500 people, operating across four countries or more. Digital transformation and buy-side M&A complete the experience we will explore; acquisitions matter here because they often bring models and data estates that were never governed to the same standard. The capabilities that count are risk judgement at level 4, then three at level 3: architecture at scale, product delivery, and data and AI governance. Your record should show a time when what the numbers said moved staff, capital or management attention elsewhere, and successors you developed while the model around them changed.

Committee chairs, executives, then the board

In the opening week the Technology Committee Chair talks with you confidentially over video. The Audit Committee Chair takes the third week, where funding discipline and model risk controls are the natural ground. Week five is a panel of executives, met in person, representing each function, and in week seven you meet the full board. We send our response in writing within three working days of each.