Interim Chief Information Officer — Scale-up

Ninety days to an AI control framework

Inside the first ninety days, the interim CIO and the Group Chief Executive Officer should agree two things: who decides what across technology, and the controls that govern AI decisions. Those controls have three parts. Evidence: what a model must demonstrate before it is trusted with a decision. Override: who can set its output aside, and how that is recorded. Retirement: when a model, rule or tool is switched off, and who is entitled to do it.

By around month nine, technology commitments should have been sorted into those that continue, those paused and those ended, and every AI system in use should have a single person accountable for it. By month 24 technology leadership and the escalation route should have scaled with the organisation, and the board should see every AI override and retirement on a regular cycle, as ordinary governance rather than special pleading.

Executive authority over technology and its AI

This is an interim executive appointment with real authority. You will set which technology decisions are taken centrally and how far local teams may go on their own authority. You will also decide what proof a technology request must bring before funding, capacity or control is granted. Each technology commitment will continue, wait or end on your call, and anything that passes cost or risk to others with no owner stops. You will withhold investment where a system's users, its upstream dependencies or its failure modes have not been thought through, settle who leads technology and how problems are raised, and work through conflicts between the chief executive and operating leaders before budgets or teams are tied up.

In a public-interest service the AI question is sharper than in most sectors. A model that misroutes a request or misjudges eligibility affects a person who often has no alternative provider. Override and retirement controls are therefore not decoration; they are how an organisation shows it is still in charge of its own decisions.

A CIO who governs models as carefully as money

You have led technology through a period of fast growth and kept governance in step with it. The businesses you have led technology for reached at least $605 million in revenue, 3,600 people and six countries, and in them the data you produced actually redirected staff, spending and leadership focus. Integration, digital transformation and scale-up experience are examined. Risk judgement is expected at level 4; at level 3 we look for product delivery, architecture that holds up at scale, and the governance of data and AI. You can tell which local discretion protects a service and which should be withdrawn, and you write that line down. You can brief the Group Chief Executive Officer on what is still unresolved and recommend a course anyway, and a technology commitment can be stopped on your word without a vote among country heads.

Selection for the CIO seat

Selection starts in week one, when the Group Chief Executive Officer talks with you confidentially by video. In the third week the Regional President spends time with you on how far AI controls can reach in practice. In week five you meet executive leaders from each function in person, and the full board of directors in the seventh week. Our written view on every stage follows within five working days.