Joseph Noujaim is a doctoral student in the Doctor of Business Administration (DBA) program at Grenoble Ecole de Management, where the research is grounded in the practical problem enterprises now face as AI systems move from supporting work to exercising delegated authority.

​

Alongside the doctoral program, the work is informed by a career spent inside large, operationally complex organizations, where governance is never abstract and where the cost of ambiguity is paid in real decisions, across real systems, under real time pressure.

​

In industry, Joseph works in senior technology leadership, with a focus on enterprise scale transformation across multiple markets, and on the design of platforms that turn fragmented operations into coherent, governable systems.

​

This practitioner vantage point is not presented as a credential for its own sake. It matters because the central question of this research, how an organization keeps its intent intact when decisions are delegated to fast, tool connected agents, is best understood at the boundary where governance meets production, where policy language encounters operational ambiguity, and where the signals of drift are felt before they are formally named.

The academic contribution being pursued is deliberately applied. The literature review is being built to connect decades of organizational control theory to the new reality of artificial agency, treating delegation, discretion, and cultural control as the baseline that enterprises already know how to manage for humans, then asking what breaks when the actor is not socialized and cannot absorb institutional sensibility through lived experience.

​

From that synthesis, the research proposes a methodology called ALiEn, short for Agency Licensing and Enforcement, designed as a set of practices that operate at runtime, not only before deployment or after incident. It aims to make mandate legible enough to travel with agents, evidence visible enough to be interpreted by governance forums, and authority adjustable enough to remain proportional as drift appears and recedes.

This combination of academic training and practitioner exposure defines the stance of the work. It is neither a product announcement nor a distant critique. It is an attempt to earn a methodology through disciplined theory, field evidence, and iterative evaluation, so that organizations can delegate to AI agents without collapsing into either blind trust or permanent manual oversight.