Deploying Artificial Intelligence Agents in Business Operations.

Avv. Alessandro del Ninno has authored the essay “Impiegare gli agenti di intelligenza artificiale nell’attività di impresa. Definizione funzionale, governance, casi d’uso e configurazione operativa” (Deploying Artificial Intelligence Agents in Business Operations. Functional Definition, Governance, Use Cases and Operational Configuration), published in Diritto Bancario on 7 May 2026.

In this contribution, Avv. del Ninno examines the deployment of AI agents within corporate processes, distinguishing them from more traditional generative AI systems. While the latter are typically used to produce, summarise or rework content, AI agents are characterised by their ability to operate within more complex procedural sequences, by collecting information, consulting documents, interacting with corporate tools and supporting the performance of operational activities.

The essay clarifies that an “AI agent” does not currently constitute an autonomous legal category under Regulation (EU) 2024/1689, but rather represents an operational mode of use of artificial intelligence systems. For this reason, its introduction within a company cannot be treated as a mere technological choice. It requires, instead, a prior assessment of the relevant processes, the data used, the information sources consulted, the permitted levels of autonomy and the internal allocation of responsibilities.

Particular attention is devoted to the organisational impact of agentic AI. Avv. del Ninno notes that such tools may affect execution times, costs, controls, output quality and the distribution of responsibilities, thereby making structured governance indispensable. From this perspective, an AI agent must be configured through clear rules, operational boundaries, logging mechanisms, escalation procedures, consistency checks and appropriate human oversight.

The essay also reviews the main corporate use cases, ranging from administrative, legal and procurement functions to complaints handling, from human resources to marketing, and through to IT and security functions. Of particular interest is the practical case concerning the use of an AI agent for the preliminary assessment of insurance claims and complaints, which enables the author to outline an operational pathway based on process mapping, definition of the agent’s mandate, construction of the documentary knowledge base, configuration of permissions, implementation of controls, testing and progressive deployment within a controlled environment.

Overall, Avv. Alessandro del Ninno’s essay offers a technically rigorous and operationally focused analysis of agentic AI, placing at its core not mere technological innovation, but the company’s ability to govern it. The underlying message is clear: AI agents may represent a powerful driver of efficiency and process quality, but only where they are embedded within an organisational framework that is transparent, auditable and consistent with the principles of compliance, accountability and risk management.

Visualizza il documento allegato