Brief IA
›

IBM: 78% of British Organizations Have an AI Director

Business & Startups·Tom Levy·

IBM: 78% of British Organizations Have an AI Director

IBM: 78% of British Organizations Have an AI Director
Key Takeaways
1According to IBM, 78% of British organizations have an AI director as of May this year, compared to 26% in 2025.
2Rhodri Arrowsmith advocates for empowering this role to suspend or redirect projects.
3He recommends specifying the actions that AI can trigger without human intervention.
Why it matters : According to Rhodri Arrowsmith, the AI director must be able to arbitrate investments and assess the outcomes of projects.

According to IBM, the proportion of British organizations with an AI director reached 78% in May of this year, compared to 26% in 2025. For Rhodri Arrowsmith, managing partner at IBM Consulting in the UK and Ireland, this appointment must come with real powers over projects as well as budgets.

The Director Must Be Able to Suspend Projects

The ability to halt an initiative, modify its scope, or redirect it: Rhodri Arrowsmith places these prerogatives at the heart of the AI director's mandate. This individual should also select the processes where technology will be deployed and, with the support of administrators, have the capacity to act across multiple components of the business. According to IBM, certain practices can promote the success of projects.

For the head of IBM Consulting, the objectives assigned to the position are as important as its authority. He also emphasizes the value of experience in distinguishing effective initiatives, failures, and untapped opportunities. In his view, how this role is defined, linked to advisory functions, and empowered will be more decisive than mere presence.

His recommendations also focus on controlling expenses and establishing precise performance criteria. Companies should quickly decide which projects to halt, which to accelerate, and which to allocate more resources to. Rhodri Arrowsmith believes that these governance and allocation choices will determine their ability to gain a competitive advantage.

Companies Must Limit the Autonomy of Systems

The control of systems constitutes another aspect of the mandate advocated by Rhodri Arrowsmith. He recommends clarifying what AI can trigger on its own and what requires human intervention. This distribution should fall under governance that makes responsibilities explicit, allows for performance comparisons against objectives, and permits intervention when necessary.

He also calls for deployments to follow standards and for the board to develop and validate the governance model. The appointment of a responsible individual should thus be part of a strategy focused on achieving concrete results, rather than serving as a control mechanism. For the next phase of transformation, which he recommends be designed and secured rigorously, the director could rely on trials already conducted and the knowledge acquired by employees.

La lettre du soirGratuit · 19 h

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

From Claims to Investments: Measuring Results

In banks as well as in retail, handling claims involves significant volumes of unstructured information. AI can analyze this data and accelerate the development of processing recommendations, potentially improving productivity and customer experience. This usage aligns with the role Rhodri Arrowsmith proposes for the director: to implement useful applications and establish their outcomes.

The head considers this period crucial for the value that companies can derive from AI, as approximately two-thirds of projects fail to meet their objectives, according to him. He recommends a common assessment based on indicators and visibility on investments. Technical architecture, standards, priorities, and funding distribution would fall within the director's scope, who is also tasked with formulating a collective vision and sharing successes.

These missions often take place in organizations that are difficult to transform. Rhodri Arrowsmith describes many of IBM's clients as large entities shaped by internal growth or acquisitions, with complex operations and technologies inherited from different eras. As this environment cannot be replaced immediately, he recommends a joint understanding of the business and tools to identify opportunities for value creation within dispersed structures.

Reporting to the CEO Remains a Prospect

The establishment of functions dedicated to AI does not come with a uniform hierarchical attachment. Rhodri Arrowsmith observes that some directors report to data teams, while others report to technology teams, which can have implications for their accountability. He believes that the CEO ultimately bears responsibility for the consequences of AI, and that a leader positioned too far from this level sees their actions limited; thus, support from the business is essential.

He envisions a role eventually situated at the board level, under the authority of the CEO or the Chief Operating Officer. He also predicts that it will become as common in executive management as the roles of Chief Data Officer or Chief Information Officer. However, its responsibilities may differ depending on the companies and the methods they deem effective.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.