The arrival of generative artificial intelligence in the public service raises as many hopes as fears.
The arrival of generative artificial intelligence in the public service raises as many hopes as fears. Too often, the debate focuses on risks: data protection, legal security, algorithmic bias or even regulatory compliance. These concerns are legitimate. But they should not obscure an essential reality: generative AI represents above all a major managerial opportunity for the public service.
Because beyond the tool, it is the organization of work, the posture of managers and collective performance that are at stake.
Concrete potential for agents and managers
In many administrations, executives devote a significant part of their time to repetitive tasks with low added value: writing notes, reports, summaries, standardized responses or even preliminary data analyses. Generative AI can act as a professional assistant, capable of producing first versions, helping to structure thinking or facilitating access to information.
Used in a controlled manner, it does not replace either human expertise or hierarchical responsibility. On the other hand, it frees up time for missions with higher added value: strategic management, team support, change management or user relations.
A cultural issue before being technological
The main obstacle to the deployment of generative AI in the public service is not the absence of tools, but the administrative culture itself. The fear of error, the logic of conformity and the verticality of organizations slow down experimentation.
However, generative AI presupposes a new managerial posture: accepting the test, regulating the use rather than prohibiting it, and trusting the professional intelligence of the agents. Without acculturation of managers, any AI strategy will remain theoretical.
Training supervisors to understand what AI does – and does not do – is therefore an essential prerequisite.
The structuring role of general management and HR departments
The responsibility of general management and HR departments is central. It is not a question of allowing informal uses to develop, but of establishing a clear framework: rules of use, authorized perimeters, traceability requirements and ethical principles.
The issue is also HR. AI questions expected skills, increased expertise, critical analysis capacity and continuing training. It requires us to rethink job standards, career paths and performance evaluation.
As such, generative AI can become a lever of attractiveness for the public service, particularly among young executives in search of meaning, autonomy and innovation.
Making AI a tool of meaning and public performance
The central question is therefore not whether generative AI should enter the public service, but how. Used without a frame, it carries risks. Ignored, it creates a disconnect with the reality of the world of work.
But conceived as a tool for assistance, professionalization and managerial modernization, it can contribute to restoring meaning to public work, strengthening the quality of the service provided and improving collective performance.
The real risk, at a time when administrations are facing major challenges, would be to miss this opportunity.
