Διοίκηση δημοσίων συμβάσεων για συστήματα τεχνητής νοημοσύνης στην πολιτική προστασία : εμπειρική ανάλυση του προγράμματος πολιτικής προστασίας ΑΙΓΙΣ και προτάσεις αξιοποίησης εργαλείων συμβάσεων καινοτομίας
Public contract management for Artificial Intelligence systems in civil protection : empirical analysis of the AIGIS civil protection program and proposals for the utilization of innovation contract tools

View/ Open
Keywords
Δημόσιες συμβάσεις ; Τεχνητή νοημοσύνη ; Πολιτική προστασία ; Δημόσιες συμβάσεις καινοτομίας ; Προ-εμπορική δημόσια σύμβαση (PCP) ; Δημόσιες συμβάσεις καινοτόμων λύσεων (PPI) ; Σύμπραξη καινοτομίας ; Τεχνικές προδιαγραφές ; Τεχνολογική αβεβαιότητα ; Διοίκηση δημοσίων συμβάσεων ; Ψηφιακός μετασχηματισμός ; Πρόγραμμα Πολιτικής Προστασίας «ΑΙΓΙΣ»Abstract
This thesis examines the management of public contracts for artificial intelligence systems in the field of civil protection, using selected subprojects of the “AIGIS” civil protection program of the Hellenic Ministry of Climate Crisis and Civil Protection as an empirical case study. The purpose of the study is to investigate the relationship between technical specifications, procurement procedures, and the evolution of tendering processes in projects characterized by high technological complexity and increased uncertainty regarding the performance and functionality of artificial intelligence systems.
The research adopts a qualitative comparative approach, combining document analysis, comparative case study methodology, and interpretive synthesis of qualitative data. Three subprojects of the “AIGIS” program concerning information systems and artificial intelligence applications in civil protection are examined. In parallel, questionnaires addressed to members of technical specification committees, tender evaluation committees, and project monitoring and acceptance committees are utilized. The analysis focuses on variables such as the clarity of technical requirements, interoperability, cybersecurity, data management, pilot testing capability, and the suitability of procurement procedures.
The findings indicate that artificial intelligence itself does not constitute a direct factor of procurement failure. Instead, critical factors include the technical maturity of the project, the clarity of operational needs, the ability to formulate verifiable technical specifications, and the existence of mechanisms for testing and gradual validation. The research further demonstrates that traditional procurement procedures present significant limitations in projects characterized by high technological uncertainty, particularly when artificial intelligence constitutes the core operational component of the system.
Based on these findings, the thesis examines the potential use of innovation procurement tools, such as Pre-Commercial Procurement (PCP), Public Procurement of Innovative Solutions (PPI), and Innovation Partnerships, as alternative or complementary mechanisms for managing technological uncertainty. Finally, an indicative decision tree is proposed to support the selection of the most appropriate procurement procedure according to the level of technological maturity and operational complexity of the project.


