Gianluca Luise*†
Giovanni Cocozza†- Department of Political Science, University of Naples Federico II, Naples, Italy
This paper explores the evolving role of AI in administrative empowerment, analyzing its implications for public administration and legal norms in Italy and the broader European context. It examines key institutional and legislative developments and the broader impact of AI on administrative law, aiming to provide a comprehensive understanding of how emerging technologies are shaping modern governance structures. Through this analysis, the study contributes to the ongoing discourse on AI regulation, emphasizing the need for a harmonized approach that fosters both technological progress and legal integrity.
1 Introduction
In recent years, artificial intelligence has emerged as a transformative force in multiple sectors, including public administration. The integration of AI in administrative processes emerges as a strategic tool to strengthen the efficiency, transparency and decision-making capacity of public apparatuses, contributing to a more responsive and accountable governance. From an administrative law perspective, AI applications can facilitate regulatory compliance, simplify processes, optimize resource allocation and promote the empowerment of administrations.
Administrative capacity thus becomes an autonomous value to be protected and strengthened, in line with the principle of good administrative performance enshrined in Article 97 of the Italian Constitution and the European Commission’s most recent indications. Strengthening this capacity is essential to counter the widespread perception of institutional inefficiency, which risks fuelling phenomena of disaffection and democratic distrust on the part of citizens.
However, while the use of AI can reduce the risk of unequal treatment, increase compliance with the law, prevent corrupt phenomena, improve investigative thoroughness and make administrative decisions more predictable and certain, it also raises complex legal and ethical issues. Issues such as accountability of automated decisions, protection of personal data and respect for fundamental rights emerge as crucial challenges to be addressed.
The European Union and Italy are among the main players in the development of regulatory frameworks aimed at regulating the use of AI in administrative procedures. The proposed European Regulation on Artificial Intelligence (AI Act), together with specific national legislative initiatives, highlights the urgency of balancing technological innovation with adequate legal guarantees in order to preserve democratic legitimacy, administrative transparency and respect for fundamental rights in the new digital ecosystem.
This study adopts a qualitative and analytical approach, based on a critical examination of legal literature, European and national regulatory sources. After an initial historical-institutional part on the conception and the role assigned to Public Administration in Italy and in the EU, in particular in relation to the processes of PA empowerment (“capacity building”), it focuses on a specific dimension of the interaction between artificial intelligence (AI) and administrative law: that relating to respect for democratic principles and the rule of law. The boundaries of the research are therefore limited to the analysis of AI applications in administrative procedures, with a focus on decision support tools, the management of bureaucratic processes and the automation of regulatory compliance, excluding broader applications of AI in the economic or social sphere.
2 Democracy and administrative capacity: a two-way relationship
Democracy represents the institutional framework within which public administration operates, defining the principles and values that guide public action. At the same time, administrative capacity constitutes the operational tool through which political decisions are translated into concrete actions, shaping their effectiveness and impact on society. This inherently two-way relationship manifests itself in two fundamental dimensions that mutually influence each other: democracy support and public policy development.
A transparent, inclusive and efficient public administration not only facilitates policy implementation but also strengthens citizens’ trust in democratic institutions (Allegretti, 2008). When citizens perceive that institutions operate fairly and effectively, the bond between state and society is strengthened, fostering the stability and legitimacy of the democratic system. Conversely, an opaque or ineffective public administration can undermine public trust, weakening the very foundations of democracy (Brairei and Alti, 2023).
On the other hand, administrative capacity plays a crucial role in determining the success of public policies. It takes the form of the set of technical, organizational and institutional skills needed to design, implement and evaluate effective interventions in response to social and economic challenges (Morisi, 2006). A highly capable public administration is not only able to solve complex problems, but can also anticipate future challenges, contributing to more informed and forward-looking decision-making.
Historically, the evolution of democratic institutions has been closely linked to the development of administrative capacities. At times of increased efficiency and innovation in public administration, democracies have shown greater resilience in the face of crises and social transformations. Conversely, periods of decline in administrative capacities have often coincided with phases of political instability and distrust in institutions.
In short, the relationship between democracy and administrative capacity is not only functional, but also historical and dynamic. Administrative capacity is not simply a tool at the service of democracy, but a constitutive element that influences its quality and durability. Likewise, democracy provides the institutional and value framework within which public administration can develop and innovate, creating a virtuous circle that strengthens both dimensions.
In a democracy, public administration (PA) is characterized by a number of distinctive features that regulate its functioning and define its role within the institutional system. Among these, political neutrality is a fundamental principle: administrative bodies must operate at the service of the community, without favoring partisan interests or specific groups. This principle ensures that administrative action is guided by the public interest and not by political or clientelistic logic (Allegretti, 2007).
Another essential pillar is the principle of legality, which requires the PA to operate in accordance with the law, the expression of the democratic will of the people. Legality not only ensures that administrative action is legitimate, but also helps to preserve the rule of law, the foundation of any modern democracy. Added to this is the principle of transparency, which encourages public scrutiny of the PA’s actions and strengthens citizens’ trust in the institutions (Merusi, 2007). Transparency, in fact, allows citizens to access information on administrative decisions and actions, promoting a relationship of trust and cooperation between the State and society.
A further key element is accountability: public officials are accountable for their actions to citizens and supervisory bodies. Mechanisms such as parliamentary controls, independent authorities and the administrative judiciary ensure that PA acts in the public interest and that any abuses or inefficiencies are sanctioned. Finally, participation is an essential principle for an inclusive democracy: PA must guarantee mechanisms that allow citizens to be involved in decision-making processes, thus strengthening the link between institutions and civil society.
However, the role of the PA is not without its tensions and criticalities. On the one hand, PA is called upon to translate the decisions taken by elected representatives into concrete actions, ensuring efficiency and speed in the implementation of policies. On the other, an overly bureaucratic administration risks creating distance between citizens and government, undermining political representativeness and fueling feelings of mistrust. This tension between administrative efficiency and democratic participation is one of the main challenges for modern democracies: finding a balance between streamlined decision-making processes and the active involvement of citizens is essential to preserve the legitimacy of institutions.
Among the main critical issues facing the PA, corruption and clientelism emerge as serious threats to democracy. Indeed, corruption undermines the principle of equality and damages the legitimacy of the PA, eroding citizens’ trust in institutions. To combat this phenomenon, measures such as anti-corruption laws and supervisory authorities are essential, but not sufficient: a culture of legality and accountability must be promoted at all levels of administration.
Another significant challenge is excessive bureaucracy, which can alienate citizens from active participation and reduce the efficiency of institutions. Targeted administrative reforms aimed at simplifying processes and improving the quality of public services can help overcome this without sacrificing democratic principles.
Finally, in times of crisis, dissatisfaction with the PA can fuel populist sentiments that call into question the very functioning of democracy. To counter this tendency, it is essential to strengthen transparency and institutional communication, promoting a constant dialogue between institutions and citizens. Only through a constant commitment to improving administrative capacity and preserving democratic principles is it possible to meet these challenges and ensure the effective and legitimate functioning of public institutions.
The issue of administrative capacity is increasingly asserting itself as a major element not only at the national level, but also in the European context. An increasing number of academic studies and empirical research conducted in recent years have shown that administrative capacity is a determining factor for the success of public policies and the stability of democratic institutions.
In this sense, improving administrative capacity is not only a technical issue, but an essential condition for the future of European integration and for strengthening democratic governance at all levels. This consensus is also reflected in the deliberations of policy-makers, who recognize administrative capacity as a crucial factor for the effective implementation of cohesion policies and the strengthening of governance at all institutional levels (Fici, 2004). Moreover, the most recent literature has shown that strong administrative capacity not only facilitates the implementation of European regulations and strategies but also helps to ensure greater legitimacy and transparency in public institutions, which are essential elements of democratic and accountable governance (Rodríguez-Pose and Di Cataldo, 2017; Domorenok et al., 2021; Surubaru, 2017; Terracciano and Graziano, 2016).
In the European context, in particular, administrative capacity plays a crucial role in the management of structural and cohesion funds, which account for a significant part of the EU budget.
With one third of the total EU budget allocated to cohesion policy programs, the management and implementation of these funds represent one of the most significant and complex tasks at European level. These programs, aimed at reducing economic, social and territorial disparities between Member States and regions, require not only substantial financial resources, but also a solid and well-structured administrative capacity. In this context, improving the governance and performance of public administrations emerges as a fundamental prerequisite to ensure effective and transparent management of EU funds (Bachtler et al., 2024).
The complexity of managing cohesion funds stems not only from the vastness of the resources involved, but also from the need to coordinate a multiplicity of actors at European, national and regional level. Without an efficient and capable public administration, the risk of inefficiencies, delays and waste becomes significant, undermining the very objectives of cohesion policy. On the contrary, a public administration with high technical, organizational and institutional skills can ensure that resources are optimally allocated, projects are implemented in a timely manner and results are rigorously monitored and evaluated.
Historically, the importance of administrative capacity in the management of EU funds became particularly evident during the enlargement phases of the European Union, when new member states with less developed administrative systems had to quickly adapt to European standards and procedures (Bache, 2008). In these contexts, support for administrative capacity building became a key element in ensuring that cohesion funds were used effectively and that development objectives were met.
Therefore, improving administrative capacity is not only a technical issue, but an essential condition for the success of European cohesion policies. Without an efficient and well-organized public administration, efforts to promote economic, social and territorial cohesion risk being thwarted, with negative consequences not only for the most disadvantaged regions, but for the entire European project. In this sense, administrative capacity is a fundamental pillar for the future of European integration, representing a bridge between political aspirations and their operational realization.
The ineffectiveness or failure of public policies cannot be attributed solely to an inadequate choice of instruments; the structural conditions that precede and influence the management of interventions must also be considered. These conditions, often overlooked in superficial analyses, represent the operational context within which policies are designed, implemented and evaluated. The concepts of institutional and administrative capacity (Polverari, 2020) relate to a set of structural characteristics that include internal management systems, monitoring and evaluation procedures, human resource competencies, and technical, organizational and infrastructural capacities. These interconnected and interdependent elements directly influence policy outcomes and the production of public goods and services, determining the quality and effectiveness of administrative action.
For example, a well-structured internal management system allows activities to be coordinated efficiently, reducing response times and optimizing the use of resources. Similarly, robust monitoring and evaluation procedures ensure that interventions are constantly monitored and adapted to emerging needs, improving their impact and sustainability. Human resource competencies, on the other hand, are a key factor for innovation and adaptation to changing contexts, while technical, organizational and infrastructural capacities provide the material and logistical basis for policy implementation.
These challenges require constant improvement of administrative capacity, which cannot be considered a static objective, but a dynamic and continuous process. Only through constant efforts to strengthen administrative structures and competences is it possible to ensure timely, correct and effective public spending. This approach is particularly relevant in the context of European cohesion policies, where the complexity of interventions and the need for coordination between different institutional levels make administrative capacity a determining factor in the success or failure of programs.
Improving administrative capacity, therefore, is not just a technical or bureaucratic issue, but an essential condition for the functioning of democratic institutions and the implementation of effective public policies. Without sustained efforts in this direction, the risk of inefficiency and failure remains high, with negative consequences for public trust and the legitimacy of institutions. In this sense, administrative capacity is a fundamental pillar for the future of modern democracies, representing a bridge between political aspirations and their operational realization.
3 The situation in Italy: a complex administrative system between lights and shadows
Italy represents a peculiar case in the European landscape, characterized by a complex administrative system with both significant strengths and weaknesses. Among the strengths, the country’s regulatory richness stands out. It has an advanced legal system capable of dealing with complex issues such as environmental protection, the organization of the health system and the guarantee of social rights. This regulatory framework, although sometimes accused of excessive bureaucratization, constitutes a solid basis for administrative action and the protection of citizens’ rights.
Another strength of the Italian administrative system is its widespread territorial presence. The articulation of public administration on a local basis, with a dense network of municipalities, provinces and regions, allows direct contact with communities and a greater ability to respond to the specific needs of territories (Torchia, 2009). However, this territorial fragmentation can also represent a challenge, as it requires effective coordination between the different levels of government to avoid duplication of competences and operational inefficiencies.
Finally, an important element is the recent push towards digitalization, with initiatives such as the Public Digital Identity System (SPID) and the development of Open Data platforms. These innovations are gradually improving the accessibility and transparency of public services, bringing public administration closer to citizens and simplifying processes that have traditionally been characterized by slowness and inefficiency. However, digitization is still at an emerging stage and requires further investment and careful regulation to ensure that its benefits are fairly distributed and that new forms of digital discrimination are not created (Aumenta et al., 2015).
Digitalization is significantly transforming the relationship between citizens and public administration, promoting greater transparency and accessibility of public services. Digital tools such as platforms for public participation and open data portals not only improve administrative efficiency, but can also strengthen deliberative democracy, fostering a more active involvement of citizens in decision-making processes (Arpaia et al., 2016). However, the adoption of these technologies requires careful regulation to avoid the creation of new forms of inequality related to access to digital resources or the ability to use them effectively.
In summary, the Italian administrative system, despite its criticalities, has elements of great potential, such as a solid regulatory base, a growing push towards digitization and a widespread territorial presence. However, to fully exploit these opportunities, constant efforts are needed to improve coordination between the different levels of government, to promote technological innovation, and to ensure that the benefits of digitization are accessible to all citizens.
The administrative reforms introduced in Italy, both at the national and local level, reflect the urgent need for organizational renewal that responds to contemporary challenges and the recommendations of the European Commission (Mény, 1999). In recent decades, Italy has faced several structural criticalities related to administrative capacity, including bureaucratic inefficiencies, fragmentation of competences and a limited culture of evaluation and monitoring. These critical issues called for a reforming intervention that would not only improve the efficiency of the public administration, but also strengthen its transparency, accountability and capacity to respond to citizens’ needs.
The recommendations of the European Commission played a key role in guiding the renewal process of the Italian administration, emphasizing the importance of a modern and well-equipped public administration to manage structural and cohesion funds. These requirements have been integrated into Italian regional development policy through the introduction of specific technical support measures in the operational programming documents (Bache, 2004). These measures aim to strengthen the technical and organizational skills of local administrations, improve monitoring and evaluation systems, and promote greater collaboration between different levels of government.
In particular, the Regional Operational Programs have foreseen specific actions for capacity building, i.e., the strengthening of administrative capacities, including staff training, modernization of IT infrastructures and the introduction of innovative project management practices. These initiatives represent an attempt to bridge the gap between the needs of European cohesion policies and the actual administrative capacities of Italian regions, which are often characterized by structural inhomogeneities and criticalities (Sumiraschi, 2017).
However, despite the progress made, the road to a fully efficient administration responsive to contemporary challenges remains long and complex. The ability to translate reforms into effective operational practices will depend not only on the availability of financial and technical resources, but also on the political will and innovative capacity of the institutions involved.
Systemic actions have been taken to improve administrative capacity, sometimes through autonomous operational programs aimed at enhancing the technical, organizational and institutional skills of local administrations. In this context, administrative innovation is not only an objective but also a tool of development policies, helping to create a more efficient, transparent and accountable public administration. In fact, a modern and well-functioning PA is fundamental for the competitiveness and economic growth of the country, since it guarantees effective management of resources and a timely response to the needs of citizens and businesses.
The significant change in the planning and management of Italian regional policy dates to the 1990s, as part of a broader process of political and institutional reform. Prior to this period, regional policy was characterized by a highly centralized approach, with the Cassa del Mezzogiorno as the main body for implementing development programs in Southern Italy. The Cassa del Mezzogiorno (Luise, 2020), established in 1950, was the main instrument of post-war development policy, but its decision-making monopoly limited the administrative capacities of the southern regions, relegating them to a passive and marginal role in the development process.
This centralized model, although initially effective in promoting major infrastructure works, has proved increasingly inadequate over time to respond to the specific needs of territories and to foster balanced and sustainable development. The lack of autonomy and administrative capacity at the regional level has contributed to creating a gap between the policies decided at the central level and their actual implementation at the local level, resulting in inefficiencies and wasted resources.
Since the 1990s, with the start of a reform process involving both the administrative and political systems, there has been a gradual decentralization of competences and resources, which has led the regions to play a more active role in the planning and management of development policies. This change represented a fundamental step in strengthening the administrative capacities of the regions, promoting a more participative and territorial approach to regional policy.
The transition to local development models represented a significant turning point in Italian regional policy, enhancing the endogenous resources of territories and strengthening the role of sub-national authorities. This approach, in line with European cohesion policies, has promoted a more participative and territorial vision of development, recognizing the importance of local specificities and the potential of each region. However, the decentralization process has encountered significant obstacles, mainly linked to the insufficient capacity of regional administrations to plan, manage complex interventions and cooperate effectively with other institutional actors.
The lack of technical and organizational skills, combined with an administrative culture that is often still tied to centralist logic, has limited the effectiveness of local development policies. In particular, the southern regions, which could have benefited most from a decentralized approach, have struggled to translate new competences into concrete actions due to structural deficiencies and a limited capacity for coordination.
To address these critical issues, specific measures were introduced in the 2000s to improve the adequacy of implementation instruments and to promote coordination between national and regional programming. These included the introduction of more advanced monitoring and evaluation systems, the training of administrative staff, and the creation of institutional networks to foster collaboration between the different levels of government. These initiatives have contributed to improving the capacity of regional administrations to manage structural and cohesion funds, but progress has been uneven and often insufficient to fill existing gaps.
Despite improvements, significant criticalities persist in the design of interventions, in the structural capacities of administrations and in the construction of effective multi-level governance. The fragmentation of competences, the lack of a shared strategic vision and the difficulty of integrating national and regional policies continue to represent significant obstacles to the success of development policies.
The decentralization process, therefore, represented an important step towards greater autonomy and empowerment of the regions, but its full implementation requires a constant commitment to strengthening administrative capacities and promoting a culture of collaboration and innovation (Onesti, 2020). Only through an integrated and multi-level approach will it be possible to overcome the existing criticalities and ensure a balanced and sustainable development of the Italian territories.
Analyses conducted at the European level show that the involvement of sub-national actors in policy implementation depends largely on the political stability and administrative capacity of regional institutions. In contexts where these institutions are weak or lack technical and organizational expertise, it is difficult to achieve positive results from European interventions, with the risk of wasting valuable resources and undermining the credibility of cohesion policies (Milio, 2001). In such situations, two main paths can be taken: invest in capacity building, i.e., strengthening the administrative capacities of the regions, or re-centralize responsibilities at the state level, reducing the role of sub-national entities in the management of funds.
4 The role of European action
Starting from the premise that administrative capacity in the management of European structural funds is crucial to ensure the effective and timely use of the resources made available by the European Union, with the aim of promoting economic and social cohesion between the different regions, the importance of sound administrative capacity has become increasingly evident in recent years. This recognition has led to an evolution at the European institutional level from a voluntary and flexible approach to a more structured and binding regulatory framework (Barbero et al., 2022).
In particular, the European Commission has introduced a number of ex-ante requirements and conditions that Member States must fulfil in order to access structural funds. These include the strengthening of administrative capacities, the adoption of advanced monitoring and evaluation systems and the promotion of effective multi-level governance. These conditions reflect the growing awareness that without an efficient and well-organized public administration, even the most ambitious policies risk failing, with negative consequences for territorial cohesion and citizens’ trust in European institutions.
In this context, capacity building has become a central element of cohesion policies, with the aim of filling existing gaps and promoting a more participatory and territorial approach to development. However, the success of these initiatives will depend on the capacity of Member States to translate European requirements into concrete actions and to adapt them to the specificities of their national and regional contexts.
Initially, administrative capacity was considered a factor in which individual member states could invest autonomously, according to their own needs and priorities. For years, the European Commission merely recommended the adoption of tools and practices to strengthen administrative capacity, without imposing specific obligations binding states to develop certain capacities in relation to the management of structural funds. The idea behind this approach was that each country, by virtue of its administrative autonomy and institutional characteristics, could identify the most appropriate ways to improve the effectiveness of the administrations involved in implementing operational programs (Marks, 1997).
However, the persistence of significant weaknesses in the management of European funds, such as delays in implementation, inefficiencies in spending, and difficulties in achieving cohesion objectives, led the European Commission to review its approach. Starting with the 2014–2020 programming cycle, increasingly binding rules were introduced to ensure that all Member States developed adequate administrative capacities. The so-called ex-ante conditionalities, for instance, established a set of requirements that states had to fulfil before accessing structural funds, including the strengthening of the technical and organizational skills of the administrations involved (European Commission, 2015).
With the 2021–2027 programming period, this development has reached an even more advanced level, through the introduction of enabling conditions. These conditions have made administrative capacity an essential regulatory prerequisite for access to European funds by defining specific criteria for each sectoral area receiving resources. Member States are now required to demonstrate not only the existence of an adequate regulatory and planning framework, but also the existence of governance systems capable of ensuring effective management, monitoring and control of European resources.
This gradual but clear-cut evolution reflects the growing realization that, without solid administrative capacity, the objectives of cohesion policies risk remaining unattainable. The European Commission has therefore taken a more active role in supporting Member States in adopting capacity-building measures, promoting the implementation of innovative governance tools and the sharing of good practices between different institutional levels.
This has led the European Commission to strongly guide Member States towards different forms of intervention, all to be pursued in parallel and supported by specific funding, to implement administrative capacity. These initiatives are articulated along three main lines, each responding to specific and complementary needs.
The first guideline is that of Administrative Capacity Building (ACB) (European Commission, 2022), which refers to the process of internal improvement of public organizations, accelerated by external actions aimed at strengthening the potential and competences of administrations. This approach focuses on aspects such as staff training, optimization of organizational processes, the introduction of innovative technologies and the strengthening of monitoring and evaluation systems. With reference to the programs supported by the cohesion funds, CBA has become a transversal principle, essential to ensure the success of all policy areas. Without adequate administrative capacity, in fact, even the best-designed policies risk failing in the implementation phase, with negative consequences for economic and social cohesion.
The second strand is Institutional Capacity Building (ICB), which focuses on strengthening national and regional reform strategies by improving coordination between the various levels of government. ICB aims to create more effective multi-level governance by promoting collaboration between central, regional and local institutions and fostering the integration of public policies. This approach is particularly relevant in the context of cohesion policies, where the complexity of interventions and the need for coordination between different institutional actors make a shared strategic vision and integrated management of resources essential.
In cohesion policy, Institutional Capacity Building (Milio, 2011) (ICB) has become a cross-cutting principle to ensure effective multi-level governance of public investment programs and to implement reform programs that increase the effectiveness of public policies. This approach promotes not only the strengthening of administrative competencies, but also the creation of coordination and collaboration mechanisms between the different levels of government, which are essential to address the complex challenges of regional development and to ensure an integrated management of European resources.
The third strand is Technical Assistance (TA), which differs from capacity building in its more specific and sectoral character. Whereas capacity building aims to strengthen administrative competencies and structures in a lasting and transversal manner, Technical Assistance focuses on the provision of specialized expertise to solve short-term, one-off needs. This form of support involves technicians and specialists intervening at specific stages of the decision-making and administrative process, providing advice, targeted training and technical solutions to address immediate critical issues or to optimize particular aspects of fund management.
Technical Assistance, therefore, plays a complementary role to the other two strands, acting as a temporary “accelerator” to overcome specific obstacles or to implement technical and organizational innovations. However, its impact can be significant, especially in contexts where local or regional administrations need immediate support to meet complex regulatory requirements or to launch particularly ambitious projects.
In summary, the three strands—Administrative Capacity Building (ACB), Institutional Capacity Building (ICB) and Technical Assistance (TA)—represent an integrated, multi-level approach to strengthen the administrative capacity of Member States. While CBA and ICB aim to create a solid and lasting basis for effective public policy management, TA provides targeted and timely support to address specific challenges. Together, these initiatives help to ensure that European resources are used efficiently and that the objectives of economic, social and territorial cohesion are achieved in an equitable and sustainable manner.
The transition, therefore, from voluntariness to explicit provision in official documents has forced Member States to change their perspective.
It should be emphasized that Italy has been urged several times, albeit in a fragmentary and disorganized manner, by the Council of the European Union and the European Commission itself, to strengthen its levels of administrative capacity and modernize its public administration.
Already in Recommendation 6 of 2012, specific recommendations on administrative reform were included in the indications addressed to Italy. In the 2013 recommendations, in view of the 2014–2020 programming, the Council highlighted that, despite the actions taken, weaknesses persisted in the efficiency of the public administration, in terms of rules and procedures, quality of governance and administrative capacity, with repercussions on the implementation of reforms and on the business environment. Therefore, Italian public institutions were requested to pursue greater administrative efficiency and improve coordination between different levels of government, as well as to adopt structural measures to better manage EU funds in the regions of southern Italy (Council of the European Union 2013, points 11, 21, and 23) (Publications Office of the European Union, 2014).
These instances were further emphasized in the Council’s recommendations in subsequent years, including those of July 2019, where it was clearly stated that the public sector’s lack of capacity, especially at the local level, to administer funding was a barrier to investment in all sectors due to complex procedures, overlapping responsibilities and inadequate civil service management. It was also pointed out that inadequate skills in the public sector limited the ability to assess, select and manage investment projects and that improving administrative capacity was a prerequisite for ensuring effectiveness in the delivery of public investment and the use of Union funds, with positive spill-over effects on private investment and GDP growth (Council of the European Union 2019, points 23 and 24).
This call to action has been taken up with vigour by the European Commission. Already in the position paper of the commission services on the preparation of the Partnership Agreement and programs in Italy for the period 2014–2020 and even more emphatically and clearly in the 2019 Country Report, Annex D, Italy was provided with guidelines for the programming of cohesion policy investments for the period 2021–2027. In the latter document, the Commission strongly emphasized that weak administrative capacity must be addressed to ensure proper implementation of regional development policy. This requires action on several fronts, such as strengthening vertical and horizontal partnership, strengthening the capacities of social partners and their participation in decision-making processes, and improving public procurement skills, thus the continuation of the Administrative Reinforcement Plans (ARPs). In the last programming period, these interventions were included in an Administrative Strengthening Strategy outlined in the Partnership Agreement. Prominent among these interventions is the creation of PRAs, drawn up on a voluntary basis for each national and regional administration responsible for managing EU programs.
In their first phase of implementation, these Plans implemented legislative and regulatory simplification, measures to streamline management and control procedures, and interventions in the field of human resources, such as recruitment of new resources, training, reorganization and performance incentives, as well as investments to improve IT infrastructure and support beneficiaries and the partnership.
The latest Country Report reiterated that further efforts are needed to improve the civil service, especially at management level, and that weak administrative capacity still limits the ability of public administrations to invest and implement policies.
Against this background, the question arises as to whether the widespread awareness of the problem of administrative capacity in Italy has been followed by concrete public policy actions to strengthen it. The answer to this question is positive. Despite persistent criticalities, several interventions to strengthen the capacity of Italian public administrations have been implemented over the years. At system level, even the latest reform of the public administration aims at simplifying and reorganizing public work, enhancing skills and renewing personnel.
As things stand, on the basis of what was envisaged and achieved with the previous “National Governance and Institutional Capacity Operational Program 2014–2020”, which mobilized around 800 million euro (between European and national resources) for the development of administrative and institutional capacity, the modernization of public administration and the strengthening of multi-level governance, the current “Capacity for Cohesion” (CapCoe) 2021–2027 NDP becomes a fundamental document.
The Program contributes to implementing the EU Council’s Recommendations to Italy for 2019 and 2020, and responds to the indications of Annex d of the European Commission’s 2019 Country Report on the functioning and efficiency of the PA. The Priorities of the NP dedicated to the MS Regions and to the North-Centre also implement the Recommendation on investing in the skills of public employees and on the efficiency and quality of local public services, while the digitization of the PA remains a cross-cutting theme of the NP, also considering the interventions envisaged by the PNRR. The development and improvement of the effectiveness of the implementation of cohesion policies are at the center of the Program’s interventions with the objective of contributing to the creation of more favorable conditions for access to rights and services by citizens in the territories where it is most urgent to reduce economic and social gaps and inequalities. In this perspective, the Program acts in coherence with the objectives of the European Pillar of Social Rights and with Policy Objective 5 “Bringing Europe closer to its citizens,” as well as in compliance with the EU Charter of Fundamental Rights and the horizontal principles set out in art. 9 of Regulation (EU) 2021/1060 (EN—PN Capacities for Cohesion AT 2021–2027).
The challenges identified and reported in the Program concern the deployment of a systemic intervention on the set of cohesion policy actors, involving the use of multiple instruments and levers for change. Indeed, the implementation framework is characterized by significant deficits in the implementation capacity of cohesion policies, particularly at the local territorial level. From the PRA experience, as mentioned above, stimulated through the 2014–2020 national and regional programming, some solutions have emerged that have put a strategic and structured approach back at the centre of regional territorial and administrative action. It sees the new Administrative Regeneration Plans (PRigAs), an evolution of the PRAs, as the most evolved way to support the Roadmap for Administrative Capacity Building promoted by the EU for the period 2021–2027. Italy’s Partnership Agreement for the 2021–2027 programming period also contains, in paragraph 9, precise indications for the rooting and development of these instruments, identifying them as essential for the achievement of the Cohesion Policy objectives.
From what has been elucidated so far, one can therefore conclude that the relationship between democracy and administrative capacity is crucial to ensure the sustainability and legitimacy of public institutions. Improving administrative capacity is an essential condition for meeting the challenges of the present and for building a PA, both at national and European level, at the service of a stronger and more inclusive democracy.
The relationship between public administration and democracy requires, however, a continuous balancing act between efficiency, transparency and participation. To strengthen this relationship, it is necessary to invest in the training and professionalization of civil servants, to promote reforms to simplify bureaucracy and, increasingly, to use technology to facilitate democratic participation and ensure transparency.
In this regard, it is crucial to stress how digitalization and new technologies, such as the use of artificial intelligence, are gradually transforming the relationship between citizens and PA, promoting greater transparency and accessibility. If, however, on the one hand these tools can strengthen deliberative democracy, on the other hand they need careful regulation to avoid digital discrimination.
5 The impact of artificial intelligence on the activity of public administration
In the Italian legal system, the digital transition process dates back in time. In fact, already in 2005, a “code” was dedicated to digital administration (legislative decree no. 82/2005, the so-called “CAD”) which brought out “digital citizenship rights” (art. 17, paragraph 1 quinquies), a category evanescent in its boundaries, but which well represents the increasingly marked revolution of the known categories. Art. 41 of the Code states that “public administrations manage administrative procedures using information and communication technologies”.
This is a process, as is known, in rapid evolution. Suffice it to say that art. 3-bis of law no. 241/1990, as innovated by the Simplification Decree no. 76/2020, states that to achieve greater efficiency in their activities, public administrations encourage the use of telematics, in internal relations, between the various administrations and between these and private individuals. Or, again, the introduction of the new “Digital Citizenship Charter” (art. 1, law 7 August 2015, no. 124) or the so-called “digital first” principle (art. 1, paragraph 1, letter b, law no. 124/2015, cit.) which make perceptible the change in social relations due to the advent of the new potential of the internet, already defined by the doctrine as “an expanded social space, unprecedented in the history of humanity […], where different subjects and phenomena mix, where roles can change rapidly and many interests find themselves in conflict”.1
But, as stated in the explanatory report on the recent Italian draft law on Artificial Intelligence,2 “artificial intelligence has a dark side that contains seeds of every kind, but also germs of every type”.
The phrase well represents the sensitivity of the issue, which, in intercepting multiple knowledge and proposing itself as a fundamental challenge of our time, prompts numerous reflections (Viola, 2018; Galetta and Corvalán, 2019; Benetazzo, 2020; Otranto, 2021; Marchetti, 2021; Di Ciommo, 2023) and generates questions, also due to the increasingly pressing need for effective “control” by humans over the phenomenon.
There is no doubt that in administrative law there are many advantages that can derive from a widespread use of artificial intelligence technologies, such as the reduction of the risk of unequal treatment; greater compliance with the law; the prevention of corruption; greater completeness of the investigation. With the consequence that the administrative decision could become more predictable with greater legal certainty.
It is true, however, on the other hand, that there is a deficit about compliance with guarantees in the administrative procedure.
Among the various critical issues that have been reported there are two profiles that appear more complex to address.
The first is that of the imputability of the decision adopted by the algorithm, which must remain in the hands of the body holding the decision-making power and, in accordance with the principle of “organic identification”, must allow for full verification of the logic and correctness of the results produced by the algorithm.
The second is the problem of full respect for private rights in the administrative procedure.
Naturally, it is the principle of transparency that suffers significantly, especially because it must necessarily be understood differently than the traditional one, while remaining relevant in the activity of the public administration. One inevitably wonders whether it is really possible to achieve full comprehensibility of the algorithmic decision and if the principle of transparency can act as a bulwark for true control of technological power.
When approaching the topic of Artificial Intelligence in public administration, the doctrine has had the opportunity to point out how it is necessary to make a preliminary distinction between deterministic algorithms (so-called “rule-based”) and self-learning algorithms (so-called “machine learning”).
The former, based on a deterministic logic, follow criteria established by the administration and, consequently, the decision-making process remains in the hands of the public administration, even if completed by the machine, as the power of choice is exercised upstream by identifying the criteria and attributing relevance to the various factors.
The latter, on the other hand, determines a shift from the human sphere to the machine itself, making it more difficult to resolve the problems related to the legality of the algorithm, since there is no human intervention in the data processing process. In this case, in fact, it is only the software that determines the structure of interests or relationships, even formally, as the intervention of the natural person official is not foreseen.
There are significant differences depending on the hypotheses considered, to compliance with the principle of transparency and to the just mentioned topic of participatory guarantees.
Thus, in the so-called “rule-based” algorithms, since it is a standardized procedure with low discretionary content, the compression of participation rights can cause fewer problems, given the relative ease in reconstructing the logical path followed by the software.
Differently, in a procedure that is developed using machine learning algorithms, the sacrifice of participation rights significantly affects the subject’s guarantee, no longer able to intervene to guide the decision and prepare the appropriate protection.
6 The principle of transparency in algorithmic decision-making
As previously stated, the difference between the different methods for adopting an automated decision has a significant impact on the principle of transparency, which has now become a fundamental element in the activity of public administrations.
Translating the technical rule into a legal rule is a difficult operation. And the difficulties are amplified when the artificial intelligence mechanisms relate to the exercise of administrative discretion.
The well-known distinction between the constrained and discretionary activity of the public administration is an element to be considered. In the first case, in fact, given the limited decision-making possibility that the public administration has under the law, the application of the algorithm causes fewer difficulties. In the second case, however, it may be verified that there is a relationship of inverse proportionality, for which the greater complexity of the discretionary decision that the artificial intelligence system must adopt corresponds to the lesser technical explainability of the functioning of the algorithm and of the logical legal process followed by the system.
The obligation to state reasons for all administrative acts certainly constitutes a guarantee of transparency. But this may not be enough, because it is difficult to provide an answer to the question of how the notions of knowability and comprehensibility of automated decisions should be interpreted. As mentioned, the technical understanding of the functioning of the software that returns the robotic decision can be very complex, as much as allowing a translation from technical to legal language. Otherwise, there is a risk of creating “a new form of technological bureaucracy, in which the algorithmic proposal is only formally validated by the official who makes the decision” (Lo Sapio, 2024), since he is in fact deprived of the ability to question the result.
Therefore, an adaptation of the administrative procedure is needed, in such a way as to reconcile obligations with the functioning methods of the algorithms, in a manner compatible with what is prescribed by Article 97 of the Italian Constitution. This also results in a different meaning of the principle of transparency, which cannot exclusively mean the explicability of the algorithmic decision, but also necessarily the translatability of the technical language on a legal level.
The same jurisprudence, in dealing with these issues, has had the opportunity to underline how “computer procedures, even when they reach their highest level of precision and even perfection, can never supplant, truly fully replacing it, the cognitive, acquisitive and judgmental activity that only an investigation entrusted to a natural person official is able to carry out and which therefore, in order to ensure compliance with the institutions of participation, procedural dialogue, acquisition of the collaborative contributions of the private individual and the interests involved in the procedure, must continue to be the dominus of the procedure itself, for this purpose dominating the same computer procedures set up as a servant and to which an instrumental and merely auxiliary role must therefore still be reserved within the administrative procedure and never dominant or surrogate of the human activity.”3
Even when the importance of a greater diffusion of automated procedures in the administrative procedure was highlighted, as they are able to “improve the quality of services provided to citizens and users” and the way was opened to the possibility of using the algorithm also for activities characterized by areas of discretion, the importance of a necessary adaptation was nevertheless underlined as it cannot be applied indiscriminately to algorithmic administrative activity “the entire law on administrative procedure, conceived in an era in which the administration was not invested by the technological revolution.” Also considering how “the fundamental need for protection posed using the so-called algorithmic IT tool is transparency”.4
However, it must be added that a connection has been found between the use of AI tools with the principles of efficiency and economy of administrative action (art. 1 law 241/90) and with the constitutional principle of good administrative action (art. 97 of the Italian Constitution), stating that “it is necessary to exploit the significant potential of the so-called digital revolution” and that “the use of computer algorithms for the assumption of decisions that concern the public and private sphere can determine a gain in terms of efficiency and neutrality.” With the consequence that algorithms could become a useful tool through which “to correct the distortions and imperfections that characterize typically the cognitive processes and choices made by human beings, highlighted especially in recent years by an impressive literature of behavioral economics and cognitive psychology. In this context, the decisions taken by the algorithm thus take on an aura of neutrality, the result of aseptic rational calculations based on data”. Significantly, however, the need to respect some guarantees is reiterated, like the “full knowability,” the “imputability” of the decision, the possibility of “control” of the process followed. In fact, the administrative judge states that the “knowability of the algorithm must be guaranteed in all aspects. This is to be able to verify that the criteria, assumptions and outcomes of the robotic procedure comply with the requirements and purposes established by law or by the administration itself upstream of such a procedure and so that the methods and rules on the basis of which it was set up are clear - and consequently subject to review”. More specifically, as determined by the Italian Council of State, knowability must be considered both with reference to the public administration that uses an algorithm, and with regard to the addressee of the algorithmic decisions (see text footnote 4).
The statement that to have a full transparency of the activity carried out, according to a different logic than the one previously followed, the technical formula must be translated into the legal rule, returns. This is also necessary to allow full scrutiny by the administrative judge.
7 Conclusions: towards a new form of administrative decision-making?
It does not seem doubtful that the changes taking place can have a significant impact on administrative decisions, which can no longer be classified within the usual dogmatic categories.
It seems also clear that the administrative power has been deeply affected and shaped by new technologies. And this is true both in the preliminary investigation phase, in procedural participation5 or with regard to the form of the act and the requirements necessary for its full effectiveness.
The evolution of public administration and regulatory framework offer interesting elements both at European and national level.
As is known, the European Parliament has, in fact, approved the so-called “AI Act”,6 which aims to offer greater guarantees of security and respect for the fundamental rights of the subjects involved. The European Union has opted for the Regulation, as happened with the GDPR for the regulation on data protection, to determine uniform and directly applicable constraints throughout the territory of the Union, with the aim of establishing a homogeneous and generally rigid regulatory framework for the Member States.
This regulatory discipline responds directly to the proposals of citizens who participated in the Conference on the Future of Europe (COFE), aimed, in particular, at strengthening the EU’s competitiveness in strategic sectors; creating a safe and trustworthy society, countering the fight against disinformation; promoting digital innovation, ensuring human oversight and the reliable and responsible use of AI, establishing safeguards and ensuring transparency; using AI and digital tools to improve citizens’ access to information, including people with disabilities.
The Regulation does not affect the competences of the States in the field of national security and uses a “risk-based” approach, classifying the impact of AI systems as “unacceptable” risk, with reference to biometric categorization and identification systems, systems that manipulate human behavior or exploit people’s vulnerabilities; “high” risk for systems that can cause harm to health, safety, fundamental rights, the environment, democracy and the rule of law (e.g., healthcare, banks, etc.), some law enforcement systems, migration and border management, justice and democratic processes (as in the case of systems used to influence elections); “minimal” risk, in the case of video games or spam filters. A system of sanctions and a series of measures to support innovation are also envisaged.
It is interesting to note that, on the one hand, the European legislator prohibits the use of deep learning systems without human support for activities classified as high-risk. However, it does not define such systems, but provides the criteria to identify them, also specifying the reference sectors. However, it remains unclear whether the administrative procedure can be included in high-risk activities.
Looking at what is also envisaged at national level, it can be noted that there is a system trend to be considered for possible further developments.
The draft law on artificial intelligence, recently approved by the Senate, confirms an approach according to which the algorithm must support the human decision. In fact, it is stated that “the use of artificial intelligence occurs in an instrumental function and in support of the provisional activity”.7
The aim of the draft law is to promote “a correct, transparent and responsible use, in an anthropocentric dimension, of artificial intelligence, aimed at seizing its opportunities”8 and improving the living conditions of citizens and social cohesion. The explanatory report on the draft law affirms the desire to pursue the objective of balancing opportunities and risks, promoting the use of new technologies, but at the same time providing solutions for risk management based on an anthropocentric vision.9
Then, art. 13 regulates the use of Artificial Intelligence in the sector of public administration activity as a tool capable of guaranteeing the good performance and efficiency of administrative activity, giving centrality to the principle of self-determination and responsibility of the person who uses it. Artificial intelligence becomes a tool for increasing the efficiency of administration; reducing the time required to define procedures; increasing the quality and quantity of services provided. Although, however, the need to ensure that interested parties know how it works and that its use is traceable is reiterated.
A system that is substantially confirmed in another regulatory provision, the relevance of which has been noted, as it is a significant innovation for the reference expressed to Artificial Intelligence, art. 30 of Legislative Decree no. 36 of 2023, containing the “Public Contracts Code.”
This provision in paragraph 1 states: “to improve efficiency, contracting authorities and granting bodies shall, where possible, automate their activities by using technological solutions, including Artificial Intelligence and distributed ledger technologies, in compliance with the specific provisions on the matter”.
Although it is a sectoral rule, it seems to have systemic relevance.
Even in this case, it is interesting to observe how the drafters of the new code have chosen to list a series of principles intended to govern the use of Artificial Intelligence systems by contracting authorities. The four fundamental principles of knowability, comprehensibility, non-exclusiveness and non-discrimination are thus stated.10
So, it seems possible to conclude that the use of artificial intelligence must be supportive of provisional activity, respecting the autonomy and decision-making power of the person who remains solely responsible for the provisions and procedures.
In this way, a principle of “non-exclusiveness of the algorithmic decision” must be affirmed, since the human contribution must be able to have the final say on the correctness or, better, on the legitimacy of the choice, precisely by checking, validating or refuting the automated decision.
The real challenge, however, consists in making the principles set forth actually practicable, to guarantee an effective space for the so-called “reserve of humanity”, that is, for a human intervention during the procedure and to give full operation, within the activity of the public administration, to the “human in the loop” model, which appears indispensable to exploit peculiar characteristics of man, such as the ability to interpret a context and make decisions guided by ethical principles (Pajno et al., 2019).
At the same time, however, avoiding incurring considerable risks, such as that of the so-called “Black Box”, which escapes any attempt at ex post explanation, maintaining unchanged the standard of intelligibility of the decisions adopted by public administrations, reconciling legal obligations, such as that of motivating administrative measures, with a decision that increasingly takes on the characteristics of a “mixed” decision.
As has been underlined, the changing form of the relationship between citizen and administration ends up constituting the precondition for a mutation of the substance of public action and the structure of the decision. The use of AI is, then, extremely useful when it enhances the natural intelligence of the human, increasing efficiency and guaranteeing.11 But, at the same time, avoiding the configuration of an “invisible Administration”, spectator of the decisions attributable to itself (D’Angelosante, 2016).
Author’s note
The first author is credited with the paragraphs: 1, 2, 3, 4 and the second author is credited with the paragraphs: 5, 6, 7.
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
GL: Writing – original draft. GC: Writing – original draft.
Funding
The author(s) declare that no financial support was received for the research and/or publication of this article.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The authors declare that no Gen AI was used in the creation of this manuscript.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Footnotes
1. ^S. Rodotà, Il diritto alla conoscenza, Concluding Report at the Seminar at the School for Booksellers held at the Cini Foundation in Venice, in www.scuolalibraiuem.it.
2. ^Draft Law n. 1146 (“Disposizioni e delega al Governo in materia di intelligenza artificiale”).
3. ^T.A.R. Lazio, Rome, sec. III bis, 10 September 2018, no. 9224.
4. ^Cons. Stato, Sec. VI, 13 December 2019, No 8472.
5. ^Consider, for example, the possibility for citizens to “participate in decision-making processes of public institutions by means of telematics”, according to art. 1, paragraph 1, lett. c of the law n. 124/2015, also using forms of prior consultation on the schemes of acts to be adopted, which recall the American model of “legislative rules” subject to “notice and comment”.
6. ^Regulation (EU) 2024/1689 of the European Parliament and oh the Council of 13 June 2024.
7. ^Article 13 (“Uso dell’intelligenza artificiale nella pubblica amministrazione”) Paragraph 2.
8. ^Art. 1.
9. ^The text consists of 26 articles, which regulate the integration of Artificial Intelligence in critical areas such as health (Art. 7) and work (Art. 10), information and confidentiality of personal data (Art. 4), economic development (Art. 5), intellectual professions (Art. 12), judicial activity (Art. 14), investments in the sector with a spending authorization of 1 billion euro (Art. 21), protection of users (Art. 23), copyright, for the specific discipline of works created with the help of IA (Art. 24) and finally criminal protection (Art. 25), with the introduction of an aggravating circumstance for offences committed by means of artificial intelligence systems, special aggravating circumstances for certain offences and the introduction of a new criminal offence.
10. ^The Paragraph 3 specifies that: “decisions taken through automation comply with the principles of: a) knowability and comprehensibility, whereby every economic operator has the right to know the existence of automated decision-making processes that concern him and, in this case, to receive significant information on the logic used; b) non-exclusivity of the algorithmic decision, whereby in any case there is a human contribution in the decision-making process capable of controlling, validating or denying the automated decision; c) algorithmic non-discrimination, whereby the owner implements adequate technical and organizational measures in order to prevent discriminatory effects against economic operators”.
11. ^It is therefore necessary to reiterate the relevance of the consideration according to which “technology that is not human-centred will not be a solution”. See Dangel et al. (2018).
References
Allegretti, U. (2007). “Democrazia partecipativa e controllo dell’amministrazione” in Democrazia e Diritto (Franco Angeli: Milano), 71–79.
Allegretti, U. (2008). “Democrazia partecipativa e processi di democratizzazione” in Democrazia e diritto, II trimestre (Roma: Edizione Tritone), 1–43.
Arpaia, C. M., Ferro, P., Giuzio, W., Ivaldi, G., and Monacelli, D.. (2016). L’e-Government in Italia: situazione attuale, problemi e prospettive, in Questioni di Economia e Finanza, Banca d’Italia, p. 27.
Aumenta, V., Lazzaroni, A., and Adda, L. (2015). L’identità digitale: dalle nuove frontiere del sistema pubblico di identificazione (spid) alle problematiche legate al web, in Ciberspazio e diritto: rivista internazionale di informatica giuridica, fascicolo. Modena: Stem Mucchi Editore. 11–28.
Bache, I. (2004). “Multi-level governance and European Union regional policy” in Multi-level Governance. eds. I. Bache and M. Flinders (Oxford: Oxford University Press), 165.
Bache, I. (2008). Europeanization and multilevel governance. Cohesion policy in the European Union and Britain. New York: Rowman & Littlefield, 39–43.
Bachtler, J., Polverari, L., Domorenok, E., and Graziano, P. (2024). Administrative capacity and EU cohesion policy: implementation performance and effectiveness. Reg. Stud. 58, 685–689. doi: 10.1080/00343404.2023.2276887
Barbero, J., Christensen, M., Conte, A., Lecca, P., Rodriguez-Pose, A., and Salotti, S. (2022). Improving the governance quality in the regions of the EU and its system-wide benefits for cohesion policy. J. Common Market Stud., 1–20. doi: 10.1111/jcms.13337
Benetazzo, C., (2020), Intelligenza artificiale e nuove forme di interazione tra cittadino e pubblica amministrazione. Available online at: www.federalismi.it, n. 16
Brairei, C., and Alti, T. (2023). “La trasparenza amministrativa come strumento di potere e di democrazia” in Rivista trimestrale di diritto pubblico, Vol. 2 (Milano: Giuffrè editore), 809–834.
D’Angelosante, M. (2016), La consistenza del modello dell’amministrazione “invisibile” nell’età della tecnificazione: dalla formazione delle decisioni alla responsabilità per le decisioni, in S. Civitarese Matteucci and L. Torchia (a cura di.), La tecnificazione, Firenze: Firenze University Press, p. 155.
Dangel, S., Hagan, M., and Williams, J. B. (2018) Designing today’s legal education for tomorrow’s lawyers: the role of legal design, technology and innovation
Di Ciommo, I. P., (2023), La prospettiva del controllo nell’era dell’Intelligenza Artificiale: alcune osservazioni sul modello Human In The Loop. Available online at: www.federalismi.it, n. 9.
Domorenok, E., Graziano, P., and Polverari, L. (2021). Introduction: policy integration and institutional capacity: theoretical, conceptual and empirical challenges. Polic. Soc. 40, 1–18. doi: 10.1080/14494035.2021.1902058
European Commission. (2015). European structural and investment funds 2014–2020: official texts and commentaries. Available online at: https://ec.europa.eu/regional_policy/information-sources/publications/legislation/2015/european-structural-and-investment-funds-2014-2020-official-texts-and-commentaries_en
European Commission (2022). Administrative capacity building self-assessment instrument. For managing authorities of EU funds under cohesion policy. Paris: OECD.
Fici, L. (2004). Governance interna, esterna e inter-istituzionale negli enti locali. Milano: Franco Angeli.
Galetta, D. U., and Corvalán, J. G., (2019), Intelligenza Artificiale per una Pubblica Amministrazione 4.0? Potenzialità, rischi e sfide della rivoluzione tecnologica in atto. Available online at: www.federalismi.it, n. 3
Lo Sapio, G. (2024). L’Artificial Intelligence Act e la prova di resistenza per la legalità algoritmica. Available online at: www.federalismi.it, no. 16, p. 281.
Luise, G. (2020). Dalla Cassa per il Mezzogiorno ai fondi europei: un’analisi istituzionale delle politiche di sviluppo per il Meridione d’Italia, pp. 245–274 in Scenari e prospettive politiche, sociali ed economiche della UE: il libro bianco sul futuro dell’Europa. Atti del XXXIII Incontro Arethuse “Gobernanza Postcrisis en la Unión Europea.
Marchetti, B. (2021). La garanzia dello human in the loop alla prova della decisione amministrativa algoritmica. BioLaw J. 2. doi: 10.15168/2284-4503-828
Marks, G. (1997). “A third Lens: comparing European integration and state building” in European integration in social and historical perspective: 1850 to the present. eds. J. Klausen and L. Tilly (Lanham: Rowman & Littlefield), 23–43.
Mény, Y. (1999). Le riforme amministrative in Italia ed in Europa, in Riforme amministrative italiane: un confronto europeo, SPISA conference, pp. 1-4.
Merusi, F. (2007). “Il principio di legalità nel diritto amministrativo che cambia” in Diritto pubblico, Rivista fondata da Andrea Orsi Battaglini (Bologna: Il Mulino), 427–444. doi: 10.1080/00343400601120213
Milio, S. (2001). Can administrative capacity explain differences in regional performances? Evidence for structural funds implementation in southern Italy. Reg. Stud. 41, 429–442.
Milio, S. (2011), Il Processo di Capacity Building per la Governance delle Politiche di Sviluppo. Il Ruolo della Capacità Amministrativa nell’Implementazione della Politica di Coesione, in Rivista giuridica del Mezzogiorno, Trimestrale della Svimez, pp. 609–658.
Morisi, M. (2006). Democrazia e governo locale. Le prospettive delle assemblee elettive. Firenze: Firenze University Press, 1–33.
Onesti, G. (2020), La capacità amministrativa delle Regioni italiane in un’ottica comparativa, in Rivista economica del Mezzogiorno, Trimestrale della Svimez, pp. 513–538.
Otranto, P., (2021), Riflessioni in tema di decisione amministrativa, intelligenza artificiale e legalità. Available online at: www.federalismi.it, n. 7
Pajno, A., Bassini, M., De Gregorio, G., Macchia, M., Patti, F. P., Pollicino, O., et al. (2019). Intelligenza Artificiale: criticità emergenti e sfide per il giurista. BioLaw J. 3:205. doi: 10.15168/2284-4503-449
Polverari, L. (2020). La capacità amministrativa: letteratura, criticità e un’agenda di ricerca a servizio del policymaking, Riv. Giur. Mezzogiorno, Trimestrale della Svimez, pp. 743–762.
Publications Office of the European Union. (2014). The European council in 2013. Printed in Belgium.
Rodríguez-Pose, A., and Di Cataldo, M. (2017). Quality of government and innovative performance in the regions of Europe. J. Econ. Geogr. 15. doi: 10.1093/jeg/lbu023
Sumiraschi, C. (2017), La sfida della programmazione comunitaria: i Programmi Operativi 2014-2020 alla prova dell’attuazione, in Riv. Giur. Mezzogiorno, Trimestrale della Svimez, pp. 53–74.
Surubaru, N. (2017). Administrative capacity or quality of political governance? EU cohesion policy in the new Europe, 2007–13. Reg. Stud. 51, 844–856. doi: 10.1080/00343404.2016.1246798
Terracciano, B., and Graziano, P. R. (2016). EU cohesion policy implementation and administrative capacities: insights from Italian regions. Reg. Fed. Studies 26, 293–320. doi: 10.1080/13597566.2016.1200033
Viola, L. (2018), L’intelligenza artificiale nel procedimento e nel processo amministrativo: lo stato dell’arte. Available online at: www.federalismi.it, n. 21
Keywords: AI, administrative empowerment, evolution of public administration, administrative decision, principle of transparency
Citation: Luise G and Cocozza G (2025) The process of “capacity building” in the public administration: artificial intelligence, transparency and new models of the administrative decision. Front. Polit. Sci. 7:1605729. doi: 10.3389/fpos.2025.1605729
Edited by:
Gema Sánchez Medero, Complutense University of Madrid, SpainReviewed by:
Ana-Maria Bercu, Alexandru Ioan Cuza University, RomaniaSonia Scognamiglio, University of Salerno, Italy
Carlo Pontorieri, University of Campania Luigi Vanvitelli, Italy
Copyright © 2025 Luise and Cocozza. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Gianluca Luise, Z2lhbmx1Y2EubHVpc2VAdW5pbmEuaXQ=
†ORCID: Gianluca Luise, orcid.org/0000-0002-3401-1926
Giovanni Cocozza, orcid.org/0009-0008-4423-7559