Orange announced on 24 June 2026 that Usman Javaid would become Group Chief AI Officer on 1 September 2026. Reporting directly to Bruno Zerbib, the group's Chief Technology & Innovation Officer, he succeeds Steve Jarrett and is tasked with accelerating and scaling Orange's AI initiatives, while the group targets more than 600 million euros of value generated by AI by 2028. The announcement raises a question any executive team can ask itself: what does a Chief AI Officer do, and do you need one?

A Chief AI Officer (CAIO) is the executive accountable for a company's artificial intelligence strategy. They set the usage doctrine, choose which use cases get funded, put governance and compliance in place, and steer the move from experiments to AI agents running in production under human control, working between the executive board, the business lines and IT.

The title spread first in the United States. At Orange, the announcement describes a role built around value creation and large-scale deployment.

What is a Chief AI Officer, and what do they actually do?

A CAIO works where the executive board, the business lines and the IT function meet, and the day-to-day work looks more like arbitration than research. Job descriptions vary in wording, yet the same responsibilities keep coming back:

An arbitration role before a technical one

None of these missions is purely technical, which shows that the AI question has left the lab and become a matter of organisation. A CAIO spends most of their time making calls: between two divisions asking for the same budget, between a spectacular use case and a modest one that can be measured, between the speed the business wants and the safeguards security asks for. That requires a right to say no that the executive committee formally recognises. Without it, the CAIO runs workshops and compiles roadmaps that nobody is bound to follow.

How the role differs from the CIO and the Chief Data Officer

The functions overlap, and that is where friction starts. The CIO is in charge of the information system, its infrastructure and its security. The Chief Data Officer, where there is one, answers for data quality, access and ownership. The CAIO owns the use of AI: which use cases reach production, on what conditions, with what level of human oversight. An agent that matches invoices needs all three, since it consumes data, plugs into the IT estate and makes decisions. The practical point to settle at appointment is therefore who signs off an agent's go-live, and who can suspend it if the agent drifts.

Why is Orange appointing a Chief AI Officer now?

The role already existed at Orange: Usman Javaid succeeds Steve Jarrett, whom the announcement credits with structuring the group's AI approach and building its technology foundations. Orange ties the timing to AI entering a new phase of large-scale deployment, which in the group's view opens transformation opportunities across all its businesses.

A mandate focused on scaling up

The mandate in the announcement is operational. Usman Javaid is expected to accelerate and scale the group's AI initiatives and to embed agentic AI across networks, customer relations, operations and services for businesses. The target of more than 600 million euros of value by 2028 comes from the Trust the future strategic plan presented in February 2026, which the group frames around trust and responsibility. On compliance the text is silent, and the job description opens with value creation.

What the reporting line says

By placing its CAIO under the executive in charge of technology and innovation, Orange keeps AI close to the people who build its networks and services and treats AI as a lever for growth and efficiency. For an operator that wants agentic AI in its networks as much as in its services, the choice makes sense. A mid-market company may have no technology and innovation division at all; in that case, we would usually recommend a reporting line to the CEO rather than to the CIO: when AI touches finance, HR and sales at the same time, the CAIO has to arbitrate between divisions over which the CIO has no authority.

A regulatory timetable that is now fixed

Regulation matters too, even though Orange's announcement does not mention it. According to the European Commission, Article 50 of the EU AI Act, which sets transparency obligations for providers and deployers of certain AI systems, has applied since 2 August 2026. Only one grace period is provided: the obligation to mark AI-generated content (Article 50(2)) only applies from 2 December 2026 to generative systems placed on the market before 2 August 2026, whereas the duty to tell people they are interacting with an AI (Article 50(1)) has no such delay. The Digital Omnibus on AI, in force since 27 July 2026, moved the obligations for stand-alone high-risk systems listed in Annex III (recruitment, credit and others) to 2 December 2027, and those for AI embedded in regulated products to 2 August 2028. Boards planning their roll-outs should read what the Digital Omnibus changed in the AI Act before fixing their own dates.

For a CAIO, the timetable has a practical consequence. A conversational agent facing customers, candidates or employees already falls, in principle, under Article 50(1), while high-risk uses such as screening job applications or assessing creditworthiness leave a little more time to document and supervise. EU AI Act compliance is therefore prepared use case by use case, starting with the agents that already deal with people.

What profile should you hire as Chief AI Officer?

The profile combines skills that rarely sit in one person. A good CAIO understands the models well enough to separate what an agent does reliably from what is still a demo, runs a project portfolio with value indicators as well as technical milestones, and knows the EU AI Act and the GDPR well enough to work with the DPO and the legal team. Above all, the role needs the credibility to defend a trade-off in front of the executive committee, including when that trade-off upsets a powerful division.

When the appointment was announced, Usman Javaid had been Chief Product and Marketing Officer at Orange Business since April 2023, leading the strategy, development and marketing of products and services as well as data and AI transformation. He previously held international roles at Vodafone from 2008 to 2019, covering areas including IoT, cloud, operations and network strategy, then leadership positions at AWS from 2019, including Cloud Services across Europe, Africa and the Middle East. A product and customer profile fits a mandate centred on value, which suggests the role is shaped first by what the company expects from AI.

Is a Chief AI Officer enough to make AI projects succeed?

No. An appointment gives the topic an owner, and then the execution work begins, as the research published since 2025 keeps showing:

Where these failures come from

These failures often have organisational causes: a badly chosen use case, inaccessible data, missing governance, or a pilot built as a demo with no sequel. Even an excellent CAIO cannot remove those causes alone, because removing them takes an organisation that absorbs the change and an execution capability that holds up in front of a board and a regulator alike.

Common mistakes after an appointment

Several mistakes come back when the role is created in a hurry:

What does a CAIO appointment change for a European company?

For a company the size of Orange, the message is one of deliberate industrialisation: AI becomes a line item with a quantified value target. For a mid-market company, the lesson to keep is that the question of AI accountability applies at any size. Who, in your organisation, arbitrates use cases? Who answers for the compliance of an agent in production? If the answer is nobody, or everybody, the problem is already there, and it will grow with every new agent.

What the Orange appointment reveals

The Orange appointment makes visible a need that any organisation deploying AI meets sooner or later: a function that holds strategy, governance and execution of AI together. The form it takes, a dedicated role, a shared responsibility or an outsourced mission, is best decided once you have written down what that function must decide.

The mandate to write before looking for a name

Before any hiring, the CAIO mandate should fit on a single page. It sets the scope (which AI systems, built in-house or bought), the decision rights (launch, refuse, suspend an agent), the budget, the reporting line and the indicators the function will be judged on. It rests on an agent registry and on escalation levels everyone knows, as set out in the LOOP™ governance protocol, so that every automated decision has an identifiable human owner.

A scenario shows why this matters. In a finance department, a bank reconciliation agent proposes matching entries every morning. The mandate states who approved its move to production, which amount triggers human validation, who receives the alert when something looks wrong and who can switch the agent off without waiting for a committee. Without those written answers, the first incident turns into a crisis meeting where everyone discovers they thought someone else was in charge.

Should you appoint a Chief AI Officer or outsource the function?

Few companies have the usage volume or the budget of a telecoms operator to justify a full-time executive seat. Between doing nothing and hiring a permanent CAIO sits a middle option: the fractional Chief AI Officer. An experienced leader takes on AI strategy and governance part-time, structures the steering, trains the internal teams and, if the company wants it, prepares an internal successor to take over the role.

What the fractional model solves

This model addresses a hiring problem: the scarcity of profiles able to speak the language of the board, of the business lines and of agent governance at once. It also avoids the token role, created to tick a box with no real mandate or means. The useful question then becomes: who, with what mandate and what resources, answers for your AI in production?

Choosing between an in-house role, a shared function and a fractional CAIO

The choice rests on criteria you can check. An in-house role makes sense when the portfolio of agents in production or in preparation is large, an AI team already exists and the company can attract this kind of profile. A function shared between IT and a business division can be enough while use cases remain few and low-risk. The fractional model suits organisations with several use cases to launch, no dedicated AI resource and an immediate governance need, for instance because an agent already handles job applications or customer files.

Koneetiv plays that role with Claude Cockpit, its outsourced Chief AI Officer offer for mid-market companies without dedicated AI resources. The Koneetiv CAIO sits on the steering committee, arbitrates between building, buying and configuring, registers every deployed agent in the LOOP™ live registry and activates new use cases at regular intervals, as a continuous service with regular deliverables.

Where to start

Start with a maturity assessment

Before creating a role or mandating an expert, you need to know where you stand. The Koneetiv AI maturity assessment places your organisation on six dimensions in a few minutes, for free, governance and industrialisation included. It shows where the main blocker sits before you assign a person or a budget to it.

Accountability to assign from the first agent

Once enterprise AI agents reach production, AI accountability has to rest with an identified person, all the more so as the obligations of the EU AI Act come into application one after another. The Chief AI Officer title remains one option among several, while the function itself has to exist before go-live, with a written mandate and an up-to-date agent registry.