63 2591-2259 / This is an open access article under the CC-BY-SA license https://creativecommons.org/licenses/by-sa/4.0/ Androniceanu, A., Colesca, S.E. (2025). Triple Helix Model and Artificial Intelligence in Public Administration. Central European Public Administration Review, 23(2), pp. 63–91 DOI: 10.17573/cepar.2025.2.03 1.01 Original scientific article Triple Helix Model and Artificial Intelligence in Public Administration Armenia Androniceanu Bucharest University of Economic Studies, Faculty of Administration and Public Management, Romania armenia.androniceanu@man.ase.ro https://orcid.org/0000-0001-7307-5597 Sofia Elena Colesca Bucharest University of Economic Studies, Faculty of Administration and Public Management, Romania sofia.colesca@man.ase.ro https://orcid.org/0000-0002-5590-0589 Received: 1. 7. 2025 Revised: 16. 10. 2025 Accepted: 29. 10. 2025 Published: 11. 11. 2025 ABSTRACT Although the Triple Helix model has been widely analysed in the con- text of innovation ecosystems, its contribution to fostering the adoption of artificial intelligence (AI) within public administration remains insuf- ficiently explored. This study addresses this research gap by examining how interactions among universities, industry, and government facilitate AI integration into digital governance across selected EU countries. Purposes: The main research objectives are to: (a) assess the digital ma- turity of the selected EU countries; (b) evaluate how Triple Helix interac- tions shape AI adoption in public administration; (c) analyse the interre- lationships among the three actors within the context of AI governance; and (d) explore the connections between each country’s AI strategy and its broader governance mechanisms. Design/Methodology/Approach: The research combines both quantita- tive and qualitative methods, utilizing data from AI Watch, the European Commission, Eurostat, Oxford Insights, and the OECD. Findings: The findings reveal significant disparities among the selected EU member states and identify critical factors that either facilitate or constrain AI integration within public administration, offering new insights into the evolving role of the Triple Helix model in the era of algorithmic governance. Practical Implications: The results are particularly relevant for public sec- tor decision-makers, researchers in governance and innovation studies, and policymakers seeking sustainable models for digital transformation and collaborative innovation. Central European Public Administration Review, Vol. 23, No. 2/2025 64 Armenia Androniceanu, Sofia Elena Colesca Originality/Value: This research presents the first cross-national empiri- cal study linking Triple Helix dynamics to AI-driven innovation in the public sector, incorporating a range of indicators. The originality of this research lies in its conceptual integration of the Triple Helix framework with the transformative capacities of artificial intelligence in reconfiguring public governance and innovation dynamics within a few EU countries. Keywords: Triple Helix model, artificial intelligence, public administration, innovation ecosystems, European Union, comparative analysis, digital transformation, digital governance. Model trojne vijačnice in umetna inteligenca v javni upravi POVZETEK Čeprav je model trojne vijačnice široko analiziran v kontekstu inovacijskih ekosistemov, je njegov prispevek k spodbujanju uvajanja umetne inteli- gence (UI) v javni upravi premalo raziskan. Ta študija zapolnjuje razisko- valno vrzel z analizo, kako interakcije med univerzami, industrijo in vlado pospešujejo vključevanje UI v digitalno upravljanje v izbranih državah EU. Nameni: glavni raziskovalni cilji so: (a) oceniti digitalno zrelost izbranih držav EU; (b) ovrednotiti, kako interakcije v okviru trojne vijačnice obli- kujejo sprejemanje UI v javni upravi; (c) analizirati medsebojna razmerja med tremi akterji v kontekstu upravljanja UI; ter (d) raziskati povezave med nacionalno strategijo za UI posamezne države in njenimi širšimi me- hanizmi upravljanja. Načrt/metodologija/pristop: raziskava združuje kvantitativne in kvali- tativne metode ter uporablja podatke AI Watch, Evropske komisije, Eu- rostata, Oxford Insights in Organizacije za gospodarsko sodelovanje in razvoj (OECD). Ugotovitve: rezultati razkrivajo pomembne razlike med izbranimi drža- vami članicami EU in opredeljujejo ključne dejavnike, ki bodisi omogočajo bodisi omejujejo vključevanje UI v javno upravo, pri čemer ponujajo nove vpoglede v razvijajočo se vlogo modela trojne vijačnice v dobi algoritmič- nega upravljanja. Praktične implikacije: rezultati so posebej relevantni za odločevalce v javnem sektorju, raziskovalce s področja upravljanja in inovacijskih študij ter za oblikovalce politik, ki iščejo trajnostne modele za digitalno preo- brazbo in sodelovalne inovacije. Izvirnost/vrednost: gre za prvo čeznacionalno empirično študijo, ki po- vezuje dinamiko trojne vijačnice z inovacijami, ki jih poganja UI, v javnem sektorju ob upoštevanju niza kazalnikov. Izvirnost raziskave je v koncep- tualni integraciji okvira trojne vijačnice s preoblikovalnimi zmožnostmi umetne inteligence pri preoblikovanju javnega upravljanja in inovacijskih dinamik v nekaterih državah EU. Ključne besede: model trojne vijačnice, umetna inteligenca, javna uprava, inovacijski ekosistemi, Evropska unija, primerjalna analiza, digitalna preobraz- ba, digitalno upravljanje. JEL: O33, O38, R53, R58. Central European Public Administration Review, Vol. 23, No. 2/2025 65 Triple Helix Model and Artificial Intelligence in Public Administration 1 Introduction In the era of the knowledge-based economy and intelligent artificial technol- ogies, innovation is no longer the exclusive result of university research or pri- vate investments, but of a strategic interaction between several institutional actors. In recent years, the European Union has placed a strong emphasis on the ethical and strategic adoption of Artificial Intelligence (AI) across public sectors. According to the European Commission’s Coordinated Plan on Artificial Intelligence (2021) and the EU AI Act (2024), member states are encouraged to integrate AI to improve administrative efficiency, transparency, and citizen- centred services. However, the level of AI adoption remains uneven across member states, with significant disparities in digital readiness, data govern- ance, and institutional capacities. The Triple Helix model is increasingly recognized as both an explanatory and operational framework for analyzing collaborative innovation processes (Etz- kowitz, 2003a, p. 298; Etzkowitz, 2003b, p. 305). Within the European Union, this approach has been embedded in various regional and national initiatives, supported by structural funds and programs such as Horizon Europe. Origi- nally proposed by Etzkowitz and Leydesdorff (1995a, p.29), the Triple Helix model offers a conceptual and practical foundation for fostering innovation through dynamic and systemic interactions among universities, industry, and government. This article analyzes the application of the Triple Helix model in different European Union member states, with a focus on good practices, public policies, governance structures, and economic and social results in the context of the accelerated growth of the integration of digital technologies and artificial intelligence (Nyathani, 2023, p. 3). Case studies from EU Member States illustrate different stages of maturity of innovation ecosystems and offer relevant lessons for future European policies (Grilli and Pedota, 2024, p. 242). The Triple Helix model is becoming increasingly relevant in the context of the adoption of artificial intelligence (AI) technologies in public administra- tion (Reis et al. 2019, p. 132). AI requires both academic expertise (universities and research institutes), applied technological solutions (industry, business), and institutional capacity for integration and regulation (government through central, regional, and local administrative authorities). Success stories from some European countries (Neumann et al., 2024, p. 121) are instructive in this regard. Estonia has developed and implemented the KrattAI project, in which the government collaborates with universities and the IT industry to create AI-based digital assistants for public administration. In Finland, the AI4Cities project brings together local authorities, universities, and companies to de- velop sustainable urban AI solutions. In France, the Paris-Saclay AI Hub project is an initiative where cutting-edge research, AI startups and public policies focused on digital ethics intersect. The Triple Helix model has proven effective in stimulating innovation in EU countries with mature infrastructure and coherent public policies (Etzkowitz, 2008, p. 149). To replicate its success across the Union, it is necessary to adapt models to the regional context, strengthen institutional capacity, and encour- Central European Public Administration Review, Vol. 23, No. 2/2025 66 Armenia Androniceanu, Sofia Elena Colesca age cross-sectoral collaboration. The future of European innovation depends on the ability of actors in the three spheres to act in an integrated and adap- tive manner in a competitive global context (Parent-Rocheleau and Parker, 2022, p. 14). The EU supports these directions through initiatives such as the Green Deal, Horizon Europe and the EIT (European Institute of Innovation and Technology), promoting responsible and sustainable innovation. The Triple Helix model provides the institutional and cultural infrastructure necessary for the integration of AI in public administration. It is not a technological mod- el, but it is becoming essential for the collaborative governance of emerging technologies, including artificial intelligence. This study focuses on a few EU member states, Estonia, Finland, Germany, France, and Romania, as representative cases that reflect varying levels of AI maturity and digital governance. Estonia and Finland are recognized for their advanced e-government infrastructures, while Germany and France illustrate large-scale administrative systems adapting to AI regulation and ethical governance. Romania represents a developing context, highlighting structural and institutional barriers to AI adoption. The comparative selec- tion included in this research enables a nuanced understanding of the Triple Helix dynamics across different levels of AI readiness within the EU. From this perspective, this research aims to identify, analyze, and categorize selected EU member states according to specific variables derived from the Triple He- lix model, within the broader context of developing and integrating diverse AI applications in public interest services. Although the principles of the Tri- ple Helix model are formally embedded in European innovation strategies, their practical implementation differs considerably across Member States. To standardize innovative performance and strengthen knowledge and tech- nology ecosystems across the European Union, a differentiated but coherent approach is needed at the level of public policies. While the literature on AI in the public sector has expanded rapidly, most studies focus on technological capabilities, ethics, or citizen trust (e.g., Janssen et al., 2020; Zuiderwijk et al., 2022). However, limited attention has been given to the institutional and collaborative mechanisms that enable or constrain AI innovation in public ad- ministration — particularly through the Triple Helix framework (interaction between government, academia, and industry). This paper seeks to bridge this gap by examining the functioning of the Triple Helix model in the con- text of AI adoption within selected EU public administrations, as well as the structural conditions that enable effective governance and innovation. The analysis conducted provides a coherent foundation of data and insights on the selected EU countries, categorized according to specific variables of the Triple Helix model, in the broader context of the rapid expansion of artificial intelligence across administrative and related domains. In the EU literature, the role of AI in public administrations is rapidly growing, being the subject of interesting empirical and conceptual studies, which are selectively pre- sented in the next section of the paper. Central European Public Administration Review, Vol. 23, No. 2/2025 67 Triple Helix Model and Artificial Intelligence in Public Administration This paper contributes to the literature, offering both theoretical insights for innovation governance and practical implications for the implementation of the EU AI Act and related digital public policies. This work is organized into four interrelated sections. The first section pro- vides a comprehensive review of key concepts and scholarly studies on the Triple Helix model, emphasizing its relevance to the application of artificial intelligence in public administrations across EU member states. The second section details the research methodology, including the study’s objectives, research questions, hypotheses, and principal variables, as well as the analyti- cal framework employed to systematically investigate the topic. The third sec- tion presents the research findings alongside a correlative analysis, highlight- ing patterns and insights derived from the data. Finally, the fourth section offers the main conclusions and underscores the study’s contributions to ad- vancing knowledge in the field of public governance and AI implementation. 2 Literature Review About the Triple Helix Model in the EU Member States The Triple Helix model, developed by Henry Etzkowitz and Loet Leydesdorff (2001, p. 24), conceptualizes the interactions between university, industry, and government as pillars of innovation in a knowledge-based economy and artificial intelligence. In the EU, the Quadruple/Quintuple Helix extensions have added the sphere of civil society and the environment, strengthening a systemic framework for digitalization and AI (Androniceanu and Georgescu, 2023; Androniceanu et al., 2022). Our research aims to analyze the synergy be- tween the Triple Helix and AI in European public administration, as reflected in the specialized literature. Henry Etzkowitz and Loet Leydesdorff (2001, p. 24, 1995b, p. 115; 1999, p. 117) laid the foundations of the model in 1995, em- phasizing the hybrid role of universities, regulated by market and governance dynamics. Leydesdorff (2009, p. 381) extends the model, emphasizing the new evolution in knowledge-based economies. Elias Carayannis and Campbell (2009, p. 221) formulated the Quadruple/Quintuple Helix, including civil soci- ety and the environment in the innovation process. Research on these topics continues at an accelerated pace in the context of the large-scale penetra- tion of artificial intelligence tools in public administration in EU countries and in other areas that provide public services (Makridakis, 2017, p. 52). Thus, an intensification of interdisciplinary approaches is observed in research in the last decade. Other researchers explore AI for the public sector, emphasizing the essential role of cross-sectoral collaboration (university–industry–govern- ment), but also the managerial and technological challenges that accompany such initiatives. Straub, Morgan, Bright and Margetts (2022, p. 162) propose an integrated framework for government AI through multiple dimensions: op- erational fitness, epistemic alignment, and normative divergence. JRC (Joint Research Centre, 2024) identified and analyzed the adoption of AI in public ad- ministration through an analysis of 574 managers based on the following main dimensions: leadership, technical-ethical skills, and governance. There is a par- Central European Public Administration Review, Vol. 23, No. 2/2025 68 Armenia Androniceanu, Sofia Elena Colesca ticular interest in analyzing the synergy of the main actors in the Triple Helix model with new applications of artificial intelligence that can greatly support the cooperation between the main actors of the Triple Helix model, as univer- sities generate knowledge, companies develop AI solutions, and the govern- ment adopts/regulates everything needed in a Triple Helix structure. Michalec et al. (2024, p. 62) show that administrative centers led by mixed teams can reduce institutional barriers. Research within the JRC has highlighted the role of AI in facilitating data interoperability through ontologies and taxonomies in European public administration. AI GOV (Straub et al., 2022, p. 163) introduces procedural, structural, and relational practices at strategic and tactical levels, which can be coordinated through Triple Helix nodes. JRC (2024) recommends developing ethical and legal skills in AI governance (Rodgers et al., 2024, p. 25). In 2023, Tangi and his team at JRC produced a large study demonstrating the added value of AI in interoperability, sup- ported by government-university-industry collaboration. The European Com- mission, through the Futurium programme (European Commission, 2023) describes concrete uses such as AI assistants for employees (HR, procure- ment, reporting), highlighting investments in governance and innovation cul- ture (Etzkowitz and Carvalho de Mello, 2004, p. 162). Consequently, there is a solid body of literature exploring the interaction between AI and public administration in the context of the Triple Helix model, from theoretical per- spectives (Ethkowitz and Leydesdorff, 1999, p. 118), to applied cases and EU frameworks (JRC, 2024). This convergence provides an integrated framework for innovation, governance and skills, but the challenge of transforming this potential into real impact on public services and democratic trust remains. The literature (Votto et al., 2021, p. 14; Etzkowitz and Klofsten, 2005, p. 246) contains detailed case study examples from various EU member tates, illus- trating how the TripleHelix model is applied in public administration, in the context of artificial intelligence (Kruger and Steyn, 2025, p. 28). For example, during the pandemic, in Spain, a group of experts coordinated by Pierra Ric- cio created a multi-disciplinary center using anonymized mobility data and public surveys for epidemiological predictions and resource allocation based on the Triple Helix model and AI applications (Riccio et al., 2022, p. 7). The Triple Helix model in Spain had as main actors the regional government, the University of Alicante and the telecom companies: Telefonica and Vodafone. Another example of the application of the Triple Helix Model is identified in Denmark in the municipality of Gladsaxe. AI applications used are the internal Chatbot “GladGPT” (ChatGPT-4), launched in 2023, for employee support. In 2017, the algorithm for detecting vulnerable families was discontinued due to transparency and bias issues. The Lüneburg district and the federal states in Germany have examples of implementing a variety of chatbots to reduce direct interaction between the administration and citizens, but the results are varied (Gill et al., 2024, p. 21; Strohmeier, 2020, p. 352). Other research- ers emphasize the importance of AI strategies at the state (Bundesländer) level: competencies, regulations, and different approaches from one state to another. Other experimental examples are identified in the Czech Repub- lic between the Ministry of the Interior and universities, using various AI ap- Central European Public Administration Review, Vol. 23, No. 2/2025 69 Triple Helix Model and Artificial Intelligence in Public Administration plications for the analysis of databases from population registers, security, and e-government. In Poland, there is a high Government AI Readiness Index (62.5%), with a progressive growth trend until 2027. The Triple Helix model consists of collaborations between the state, research institutions, and pri- vate companies specialized in ICT, focusing on chatbots, predictive analysis, and smart cities. Applied research and successful implementations of the Triple Helix model also exist in Norway in the municipality of Trondheim, where a study was conducted in approximately 200 public institutions, and 19 interviews were organized. The study analyzed the early adoption of AI, with risks of discrimination and political pressure. In Italy, the PRISMA project was carried out (2016): an interoperable cloud platform for citizen engage- ment (Catania and Siracusa municipalities) - Triple Helix models between lo- cal governments, universities, and the IT industry. AIDA is an AI system based on deep learning for crime prevention and collaboration between academia, police, and the public sector. In Belgium and Latvia, AI innovations in adminis- tration based on the Triple Helix model were carried out. In Belgium, several experimental research studies were conducted, and projects were imple- mented. One of the successful research areas was Knowledge management with generative AI (smart regulation). Another was Job matching (Jobnet). As can be seen, the Triple Helix Model in the context generated by AI in public administration in the public sector in EU countries is confirmed, but the risks of discrimination and cultural barriers represent major challenges (Ranga and Etzkowitz, 2013, p. 247). Theoretical frameworks (Nosratabadi et al., 2019, p. 18; Rodgers et al., 2024, p. 27.) and studies from different countries (Es- tonia, Italy, Germany, Spain, etc.) show that success comes from ethical poli- cies, human involvement, and adaptive organizational culture, but also from massive investments in infrastructure and in the training of human resources and citizens culture (Lorincova et al., 2024, p. 19; Michalec, 2024, p. 64). Dvor- ský (2025, p. 98) emphasizes the importance of using AI in risk management. Expert recommendations include piloted approaches, rigorous training, and transparency and accountability mechanisms (Stachová et al., 2024). AI trans- forms HR in the public sector – from recruitment, performance appraisal, to professional development and planning, emphasizing the efficiency and personalization of new models, but warns of the risks of bias, transparency, and confidentiality (Mwita and Kitole, 2025). Nosratabadi et al., 2022, p. 23) present a systematic approach (“Employee Lifecycle Management”), high- lighting the use of algorithms (Random Forest, SVM, Neural Networks) at all stages of the employee life cycle – recruitment, retention, and off-boarding. Rodgers et al. (2023) address “ethical decision-making” in HR, emphasizing the integration of responsible AI to eliminate bias. The Springer (2025) study from Tanzania shows that the success of AI implementation in HR requires technological infrastructure and organizational capital. The study draws at- tention to organizational culture and the fact that AI acceptance requires managerial involvement, cultural alignment and dedicated task forces. Through the #KrattAI initiative, Estonia is developing virtual robots to guide staff and citizens. The “job matching” system has been extended to public HR, suggesting positions in line with users’ profiles. The Jobnet system uses Central European Public Administration Review, Vol. 23, No. 2/2025 70 Armenia Androniceanu, Sofia Elena Colesca machine learning to match candidates to open positions in public administra- tion (Garg et al., 2022, p. 607). A JRC comparative study conducted in Italy and Germany (N=1411) shows that human supervision does not prevent dis- crimination in AI decisions. A “fair AI” can reduce gender bias, but remains influenced by the prejudices of the assessor. Discrimination remains a prob- lem: even algorithms designed for fairness can be influenced by the actions of human assessors (Meshram, 2023, p. 329; Androniceanu, M., 2024, p. 91; Androniceanu, M., 2025, p. 110). Rigid or bureaucratic institutional culture re- duces AI adoption. Solutions: task force, “skunk works” require training and early involvement of institutional leadership (Malin et al., 2023, p. 8). Adopting AI requires clear rules, transparency, and trained personnel who can override automated decisions, as recommended by various researchers (Alaa, 2023, p. 348; De Alwis et al., 2022, p.190; Rahman and Audin, 2020, p. 259). Relevant empirical studies to discover the implications of the Triple Helix model and AI applications are identified in the literature (Wirtz, 2018, p. 609). Some of the relevant ones are presented below. In Sweden, an empirical analysis was carried out in public administration with HR analytics. The research that was carried out was based on national, regional, and local data obtained through semi-structured interviews for 51 respondents, all middle-level managers in the Swedish public sector. As the results presented in the paper “Reasons for HR analytics adoption in public sector organisations” show, organisations cur- rently use only descriptive HR analytics (dashboards, reports), but intend to evolve towards predictive analytics. The identified determining factors were: internal pressure for efficiency, the need to quantify HR indicators, and data availability (Androniceanu, 2025, p. 82; Androniceanu, 2024, p. 110). The main challenges are: limited technological capacity, lack of skills, and con- servative organisational culture. The research showed that success depends on data infrastructure, technical skills, and strategic direction, including the qualitative involvement of HR and non-HR actors. The main advantages and challenges of the penetration of artificial intelligence in human resource man- agement in public institutions, identified by Mwita and Kitole in the work pub- lished in 2025, are presented in Table 1. The findings reported by Mwita and Kitole (2025, p. 12) highlight the poten- tial risks of decision-making errors associated with the use of automated arti- ficial intelligence systems in human resource management within public insti- tutions, as illustrated in Figure 1. Central European Public Administration Review, Vol. 23, No. 2/2025 71 Triple Helix Model and Artificial Intelligence in Public Administration Table 1: Potential benefits and challenges of artificial intelligence in human resource management in public institutions HRM Component AI Benefits Strongly Agree Agree Neutral Disagree Strongly Disagree Workforce Planning and Recruitment AI automates candidate screening to identify the best fit efficiently 87 (40%) 65 (30%) 43 (20%) 17 (8%) 5 (2%) AI enhances onboarding with personalized training modules for new hires 85 (39%) 67 (31%) 42 (19%) 18 (8%) 5 (2%) Performance Management AI generates actionable insights through continuous performance tracking 78 (36%) 70 (32%) 43 (20%) 19 (9%) 7 (3%) AI improves feedback accuracy with unbiased data analysis 80 (37%) 68 (31%) 44 (20%) 18 (8%) 7 (3%) Training and Development AI customizes employee training to address individual skill gaps 81 (37%) 74 (34%) 41 (19%) 15 (7%) 6 (3%) AI identifies future skill requirements for strategic upskilling 83 (38%) 70 (32%) 43 (20%) 15 (7%) 6 (3%) Compensation and Benefits AI uses predictive analytics to design competitive salary structures 83 (38%) 72 (33%) 39 (18%) 17 (8%) 6 (3%) AI simplifies benefit administration by automating complex processes 80 (37%) 73 (34%) 41 (19%) 17 (8%) 6 (3%) Employee Relations AI monitors employee sentiment to detect early signs of dissatisfaction 74 (34%) 69 (32%) 50 (23%) 19 (9%) 5 (2%) AI recommends proactive conflict resolution strategies 76 (35%) 67 (31%) 51 (24%) 18 (8%) 5 (2%) Compliance and Legal Framework AI ensures adherence to regulatory changes with real-time alerts 76 (35%) 73 (34%) 42 (19%) 20 (9%) 6 (3%) AI reduces human errors in compliance documentation and auditing 78 (36%) 71 (33%) 43 (20%) 19 (9%) 6 (3%) Central European Public Administration Review, Vol. 23, No. 2/2025 72 Armenia Androniceanu, Sofia Elena Colesca HRM Component AI Benefits Strongly Agree Agree Neutral Disagree Strongly Disagree Workplace Health and Safety AI tracks workplace conditions to predict and prevent safety risks 79 (36%) 68 (31%) 44 (20%) 19 (9%) 7 (3%) AI facilitates wellness programs by monitoring employee health metrics 81 (37%) 69 (32%) 43 (20%) 17 (8%) 7 (3%) HR Information Systems (HRIS) AI streamlines data management and automates repetitive tasks in HR 85 (39%) 71 (33%) 41 (19%) 14 (6%) 6 (3%) AI provides predictive analytics for strategic workforce planning 82 (38%) 72 (33%) 42 (19%) 15 (7%) 6 (3%) Succession Planning AI identifies potential leaders through performance data analysis 84 (39%) 69 (32%) 44 (20%) 16 (7%) 4 (2%) AI maps career pathways to ensure seamless succession transitions 83 (38%) 68 (31%) 45 (21%) 17 (8%) 4 (2%) Employee Engagement and Retention AI predicts attrition risks by analysing engagement trends 82 (38%) 72 (33%) 43 (20%) 16 (7%) 4 (2%) AI designs personalized retention strategies using employee data insights 83 (38%) 71 (33%) 44 (20%) 15 (7%) 4 (2%) Source: Mwita, K. M. and Kitole, F. A. (2025). Potential benefits and challenges of artificial intelligence in human resource management in public institutions. Discover Global Society, 3(35), pp. 1–19. Central European Public Administration Review, Vol. 23, No. 2/2025 73 Triple Helix Model and Artificial Intelligence in Public Administration Figure 1: Risks Associated with AI in Human Resource Management Source: Mwita, K. M., Kitole, F. A. (2025). Potential benefits and challenges of artificial intelligence in human resource management in public institutions. Discover Global Society, 3(35), pp. 1–19. Another example of relevant empirical research conducted in the Swedish public administration is about AI systems integrated in the Swedish Public Employment Service (PES). The study was conducted in 2024 and the main re- sults were published in the same year (Berman et al., 2024, p. 9). The theoreti- cal framework of the research was based on Institutional Theory, Resource- Based View, and Ambidexterity. Case analysis was used as the method, with an emphasis on transparency, interpretability and stakeholder involvement. The results show that AI is effective in supporting decisions (assessment of assistance applicants), but transparency and stakeholder participation are suboptimal, which requires constant audits. Another study was conducted in 2021 in Germany by Kern and his team (Kern et al., 2021). The study was con- ducted on the German employment service, and the analysis of administra- tive data for profiling the unemployed. The methodology consisted of a com- parative evaluation of predictive models with a focus on accuracy and equity. The results of the study highlighted the existence of performing models, but differentiated in the classification of policies, which can generate inequities, requiring a rigorous audit before the implementation of the system. Inter- esting experimental research was conducted in the Netherlands (Alon Barkat and Busuioc, 2021, p. 539) on the interaction between human resources and AI in public administration. Extended and diversified results were published three years later (Alon-Barkat and Busuioc, 2023, p. 157). The methods used were controlled psychological experiments with public decision-makers. The Triple Helix model provides a solid theoretical framework for understanding AI-based public innovation ecosystems in public administration. However, to translate the potential into real outcomes – such as improved decisions and citizen trust – robust research, cross-sectoral skills, and transparent and ac- countable governance are needed. The research underlying this paper con- Central European Public Administration Review, Vol. 23, No. 2/2025 74 Armenia Androniceanu, Sofia Elena Colesca tributes to a classification of the selected EU countries and to an analysis of the state of implementation of the three pillars of the Triple Helix model in EU governance processes, in the context of artificial intelligence. In summary, the existing literature demonstrates that the integration of the Triple Helix model with artificial intelligence in public administration remains an emerging yet up-and-coming field. While substantial progress has been made in conceptualizing collaborative frameworks among government, ac- ademia, and industry, empirical evidence on effective implementation and long-term outcomes is still limited. Cross-country analyses, particularly within the EU, reveal varying levels of maturity in AI governance and innovation eco- systems. Consequently, further research is needed to explore how the Triple Helix model can foster responsible, transparent, and ethically grounded AI adoption in public governance, contributing to both administrative efficiency and democratic accountability. 3 Research Methodology According to the specialized literature, the variables for the Triple Helix mod- el are specific to the three main actors: the academic environment provid- ing knowledge, research and innovation, the business environment (industry) providing smart solutions and technologies in the context of public-private partnerships and the government/public administration, which develops public policies, provides total or partial financing and is the beneficiary of re- search and innovation. This study adopts a comparative, mixed-method re- search design to explore the relationship between Artificial Intelligence (AI) adoption in public administration and the Triple Helix model of innovation across selected EU member states. The study focuses on a few EU member states. This design enables comparative insights rather than exhaustive cov- erage, aligning with the study’s aim to analyze structural and institutional de- terminants of AI adoption in public administration. This purposive sampling enables a comparative analysis of the Triple Helix dynamics and institutional factors influencing AI adoption in different administrative and policy environ- ments, rather than aiming for exhaustive coverage of all EU states. The study employs a purposive sampling approach. A few EU countries were selected to reflect a diversity of AI maturity levels, administrative capacities, and geo- graphical contexts within the European Union. The selection was based on three main criteria: (1) diversity in AI maturity and digital governance levels; (2) institutional and geographical representation of both Western/Northern and Central/Eastern Europe; and (3) availability and comparability of data from official EU and OECD databases. This balanced selection allows for cross- country comparison of institutional drivers and constraints affecting AI adop- tion in public administration. Quantitative indicators were compiled, corresponding to the implementation of the EU Coordinated Plan on Artificial Intelligence (2021 Review) and the Digi- tal Europe Programme. Data were obtained from Eurostat, the OECD, and the European Commission’s Digital Economy and Society Index (DESI) and Oxford Central European Public Administration Review, Vol. 23, No. 2/2025 75 Triple Helix Model and Artificial Intelligence in Public Administration Insights. To ensure cross-country comparability, all variables were normalized using min–max scaling (0–1 range). Missing data (less than 5% of the total) were addressed through linear interpolation based on available time series. The empirical analysis proceeds in descriptive statistics and Pearson correla- tion coefficients were used to identify preliminary associations between AI adoption and the Triple Helix dimensions. All variables were standardized (z- scores), and multicollinearity was tested using the Variance Inflation Factor (VIF < 3). Model assumptions regarding normality, homoscedasticity, and in- dependence were verified before interpretation. This transparent methodo- logical framework ensures replicability and provides a robust foundation for interpreting the institutional mechanisms underlying AI adoption in selected EU public administrations through the lens of the Triple Helix model. Within the framework of this research, the following aspects are analyzed: (1) the degree of digital maturity of the selected EU states; (2) the intensity of the Triple Helix model; (3) the proportion of ICT graduates and the degree of in- tegration of AI in public administration; (4) collaborative framework between university, industry and government and the AI integration in public adminis- tration; (5) ITC infrastructure supporting the TH model and the integration of AI in public administration (6) the national AI and governance strategy. The objectives of the research are: (1) to analyse the degree of digital ma- turity of the states included in the pilot study; (2) to identify the impact of the Triple Helix Model (university-industry-government collaboration) on the assimilation and integration of AI in public administration; (3) to identify the main correlations between the three actors in the context of artificial intel- ligence; (4) to discover the main correlations between the TH maturity and the ICT infrastructure; (5) to find out the relations between the AI national strategies, government strategies and the degree of AI implementation in the public administration of the selected EU state. The main questions answered by the research are: (1) What is the relationship between the intensity of the Triple Helix collaboration and the degree of digi- talization and integration of AI in public administration? (2) Are there signifi- cant differences between the selected EU states in terms of implementing AI strategies in the public sector? (3) Is there a collaborative frame for supporting innovation and AI integration? How do the national AI strategies of the selected EU countries influence the success of AI implementation in administration? The main research hypotheses (H) are the following: H1. A country’s digital maturity level (DESI) mediates the relationship betwe- en the Triple Helix and the success of AI implementation in public admini- stration. H2. Countries with strong Triple Helix (TH) have a higher intensity of UIG (uni- versities, innovation, and government) collaborations. H3. Countries with higher proportions of ICT graduates tend to report grea- ter levels of AI adoption. Central European Public Administration Review, Vol. 23, No. 2/2025 76 Armenia Androniceanu, Sofia Elena Colesca H4. A well-developed collaborative framework between universities, indust- ry, and government is closely associated with increased innovation activi- ty and AI integration. H5. There is a positive correlation between TH maturity, innovation activity and infrastructure. H6. Even if states have an artificial intelligence strategy for public administra- tion, the real impact is different. The main quantitative variables used in the first stage of the research are: (1) Number of AI projects in public administration; (2) DESI index – digital public services; (3) Percentage of civil servants trained in AI; (4) Percentage of AI integration in public administration; (5) Existence of a public AI strategy; (6) Triple Helix maturity level. A special category of subsidiary research variables was added to uncover the relationships between the national AI strategies of the selected countries and their governance. This subcategory of variables includes: the level of integration of AI in the governance process; public in- vestments and partnerships with private sector organizations for the imple- mentation of AI applications in the governance process; and the degree of adoption and integration of AI in the public sector. The main research variables used for the comparative analysis are presented in Table 2. Table 2: The main research variables Research variables Type Measurement level Number of AI projects in the administration, public-private partnership Quantitative Absolute (number) Index – services publicly available digital Quantitative Index (0-100) Percent officially trained in AI Quantitative Percent Percentage of integration of AI in public administration Quantitative Percent The existence of a public AI strategy Qualitative Nominal (Yes/No) Triple Helix Maturity Level Qualitative Ordinal (low/medium/ high) The research model is represented in Figure 2. Central European Public Administration Review, Vol. 23, No. 2/2025 77 Triple Helix Model and Artificial Intelligence in Public Administration Figure 2: The correlational research model Triple Helix (Universities- Innovation-Governments) Innovative capacity Degree of assimilation and integration of AI in administration Digital maturity (moderating variable) Source: authors. 4 Discussions of the Research Results The first parameter analyzed is the digital maturity of the selected EU states, which is measured with DESI. This index includes four relevant sub-dimen- sions: (1) human capital, which measures the digital skills of the population, including the number of IT specialists and the degree of internet use; (2) con- nectivity, which measures digital infrastructure, such as broadband network coverage and internet speed; (3) Digital technology integration, which meas- ures how companies integrate digital technologies into their activities, includ- ing e-commerce; and (4) Digital public services, which measures the degree of digitalization of public services, such as e-government. The analysis combines descriptive statistics and correlation analysis to exam- ine the relationship between AI adoption in public administration and the three dimensions of the Triple Helix model. Correlation coefficients were first used to identify bivariate relationships among the indicators. Based on the data presented in Table 3, the Pearson correlation between DESI, the global DESI indicator for the year 2024, and the Government AI Readiness Index (Ox- ford Insights, 2024) for a few EU countries was determined. Central European Public Administration Review, Vol. 23, No. 2/2025 78 Armenia Androniceanu, Sofia Elena Colesca Table 3: DESI and the AI Score of the selected EU member states Country Score DESI (%) Score AI (%) Estonia 91.0 72.0 Finland 89.0 70.0 France 86.0 68.0 Germany 85.0 65.0 Romania 65.0 50.0 Sources: European Union, 2022 and Oxford Insights, 2023 The Pearson correlation coefficient calculation between DESI (European Union, 2022) and AI scores for the five countries is 0.98, indicating a very strong and positive correlation between these indicators. The correlation coefficient of 0.98 suggests that countries with higher DESI scores also tend to have higher AI scores, indicating a strong link between overall digitaliza- tion and the use of AI technologies. Estonia stands out with high scores on both DESI and AI, being a successful example of integrating digital tech- nologies and AI. Romania has lower scores on both indicators, suggesting the need for more effective investments and public policies in the field of digitalization and AI development. The comparative analysis reinforces the fact that digital maturity (DESI) and AI in public administration are funda- mental in mediating the effects of the Triple Helix on the assimilation and integration of AI in the public sector. The results show a clear positive cor- relation: countries with a higher DESI index have a significantly higher AI adoption. These results confirm hypothesis 1. Countries with solid ecosys- tems in these areas position themselves at the forefront of the EU, and the others have major potential if they invest strategically. Countries with a low DESI (Romania) have weak AI adoption. The bar chart in Figure 3 illustrates a comparative analysis of the level of AI implementation in public administra- tion and the maturity of the Triple Helix model—representing university-in- dustry-government collaboration—across five European countries: Estonia, Finland, France, Germany, and Romania. The data reveal notable differences in how these two dimensions correlate within each national context. Estonia emerges as a frontrunner in terms of AI implementation, achieving the high- est possible score, while its Triple Helix maturity is slightly lower (4). This suggests that the country’s advanced digital governance infrastructure and strong political commitment to innovation may compensate for a moder- ately developed collaborative ecosystem. Central European Public Administration Review, Vol. 23, No. 2/2025 79 Triple Helix Model and Artificial Intelligence in Public Administration Figure 3: The relationships between the AI implementation and the Triple Helix maturity Source: authors. In contrast, Finland and Germany display a reverse trend: both countries have achieved maximum maturity in their Triple Helix systems (5), but their AI im- plementation in administration remains at 4. This may indicate that despite robust collaboration among universities, industries, and governments, trans- lating such systemic strengths into administrative AI adoption requires ad- ditional strategic alignment or regulatory innovation. France demonstrates equilibrium, with identical scores (4) for both AI implementation and Triple Helix maturity, reflecting a balanced relationship between ecosystem collab- oration and technological integration. Romania, however, lags significantly behind, scoring only 2 in both categories. This parallel low performance high- lights systemic weaknesses in both administrative innovation and collabora- tive capacity, suggesting that foundational reforms are necessary to foster both ecosystem maturity and technological uptake. Overall, the comparison highlights a complex interplay between AI adoption and the maturity of collaborative ecosystems. While a high level of Triple Helix maturity appears conducive to technological implementation in some coun- tries, it does not guarantee rapid AI integration in public administration. Con- versely, Estonia’s performance suggests that strong political will and a digital- first strategy can accelerate AI implementation even when ecosystem maturity is relatively moderate. These findings underscore the need for nuanced policy approaches tailored to each country’s unique innovation landscape. The second parameter is the Triple Helix (TH) intensity – proxies. It is deter- mined based on ICT graduates and partnerships between public institutions and private sector organizations. At the level of the EU and according to DESI, Estonia leads in % ICT graduates (~11%) by 2022, while Italy, Belgium, and Cyprus have a process of less than 3%. The DESI–DII comparison shows that Germany, Finland, and the Netherlands have consolidated multi-sector eco- systems. We find out that countries with strong TH (e.g., Finland, the Nether- Central European Public Administration Review, Vol. 23, No. 2/2025 80 Armenia Androniceanu, Sofia Elena Colesca lands, Germany) have a higher intensity of UIG (universities, innovation, and government) collaborations. These results partially confirm Hypothesis 2. The graph in Figure 4 (% ICT graduates vs. AI adoption) confirms a moderate correlation: more ICT graduates usually mean a higher capacity to implement AI. However, there are exceptions (Spain has high AI adoption even with an average ICT percentage). Figure 4: The correlation between the ICT graduates and the AI implementation within the sampled member states Source: authors. The scatterplot illustrates the correlation between the percentage of ICT graduates and the level of AI adoption across selected European countries. A positive relationship is evident, as indicated by the upward-sloping trendline: countries with higher proportions of ICT graduates tend to report greater lev- els of AI adoption. These results confirm hypothesis 3. Finland and Denmark stand out as clear leaders. Finland, with over 10% ICT graduates, achieves an AI adoption rate of approximately 18%, while Den- mark surpasses 22% AI adoption with around 8% ICT graduates. These outli- ers suggest that while a strong ICT educational base is critical, other factors such as innovation policies, RandD investment, and digital infrastructure also play significant roles in facilitating AI integration. In contrast, Romania and Poland are positioned in the lower left quadrant of the chart, reflecting both a limited percentage of ICT graduates (below 4%) and low AI adoption rates (below 5%). This clustering suggests structural challenges in developing a skilled workforce and integrating advanced technologies. Central European Public Administration Review, Vol. 23, No. 2/2025 81 Triple Helix Model and Artificial Intelligence in Public Administration Germany’s performance is noteworthy: with a relatively moderate proportion of ICT graduates (about 5%), it achieves a comparatively high AI adoption rate (~15%). This deviation from the trendline indicates the potential influence of non-educational factors such as industrial capacity, public-private partner- ships, and strong innovation ecosystems. Overall, while the data support the hypothesis of a positive correlation be- tween ICT education and AI uptake, the dispersion of data points around the trendline implies that ICT graduate rates alone are insufficient predictors of AI adoption. A comprehensive approach encompassing education, policy, in- frastructure, and ecosystem development is likely necessary to drive AI inte- gration at scale. The third parameter is the degree of assimilation and integration of AI in public administrations in selected EU countries. According to AIWatch data (JRC, 2022, p. 126), in most EU countries, AI is used for multi-sector collabora- tions, governance processes, and interoperability. High levels of integration and use of AI applications in the public sector are in Denmark (24%), Portugal (17%), Finland (16%), in contrast to Romania/Poland/Hungary (~3%). These results indicate a strong correlation between digital maturity, the intensity of Triple Helix collaborations, and AI integration. The scatterplot presented in Figure 5 illustrates the relationship between the number of public-private AI projects, the presence of innovation hubs, and the maturity of the innova- tion ecosystem, measured through the Triple Helix model within a few se- lected EU countries. Figure 5: The main relationships between the research variables Source: authors. Germany emerges as the most advanced ecosystem, with the highest number of AI projects (15), the strongest presence of innovation hubs (7), and the highest Triple Helix maturity (5). This suggests that a well-developed collabo- rative framework between universities, industry, and government is closely associated with increased innovation activity and AI integration. These results conform to Hypothesis 4. Central European Public Administration Review, Vol. 23, No. 2/2025 82 Armenia Androniceanu, Sofia Elena Colesca Countries with clear artificial intelligence strategies in the public sector have higher adherence to government AI projects and significant budget alloca- tions. France and Finland occupy intermediate positions, with a moderate number of AI projects (9–10) and innovation hubs (4–5), alongside relatively high ecosystem maturity (4–5). Estonia, despite having a comparable number of AI projects (12), demonstrates a lower presence of innovation hubs (3), which may limit the diffusion and scaling of innovation outcomes beyond spe- cific sectors. Romania is an outlier, characterized by the lowest values on all three dimensions: minimal public-private AI projects (3), limited presence of innovation hubs (1), and low ecosystem maturity (2). This underscores signifi- cant structural gaps that could hinder its capacity to leverage AI for socio-eco- nomic development. Overall, the data suggest a positive correlation between Triple Helix maturity and both innovation activity and infrastructure. These results conform to Hypothesis 5. The fourth parameter analyzed is national AI strategies and governance. For this purpose, in Table 4, relevant data for the 6 countries were collected based on the most recent official reports and EU sources (AI Watch, Oxford Insights, European Commission and OECD). Table 4: The key research variable for measuring the relationship between AI governmental strategy and governance Country AI Strategy (Y/N) Governance Level Public Investment (€/capita) Ethical Regulation Public Sector AI Adoption and Integration Finland 100 90 80 90 70 Denmark 100 85 75 85 65 Netherlands 100 80 85 80 75 Germany 100 95 90 85 80 Poland 80 65 50 55 45 Romania 70 60 40 50 35 Notes: Values are normalized based on recent reporting and estimates to allow for clear comparisons. Sources: European Commission, 2022a,b; Oxford Insights, 2022 and OECD, 2022 Analysis of AI Strategies and Governance in selected EU countries is presented in Figure 6 based on a representative research variable. This radar chart pro- vides a comparative overview of six EU member states—Finland, Denmark, the Netherlands, Germany, Poland, and Romania—regarding their Artificial Intelligence (AI) strategies and governance. Central European Public Administration Review, Vol. 23, No. 2/2025 83 Triple Helix Model and Artificial Intelligence in Public Administration Figure 6: The relationship between AI Strategy and governance Source: authors. The chart visualizes five key indicators: the existence of a national AI strat- egy, governance level, public investment per capita, ethical regulation, and AI adoption within the public sector. The key findings are presented and ana- lyzed below. Finland, Denmark, the Netherlands, and Germany show a strong presence of national AI strategies (all scored 100), reflecting early and com- prehensive government commitment to AI development. These countries also score highly in governance and ethical regulation, indicating robust in- stitutional frameworks and alignment with EU ethical guidelines. Germany leads in public investment per capita, followed closely by the Netherlands and Finland, suggesting prioritization of AI research and infrastructure. Po- land and Romania exhibit considerably lower investment levels, highlighting disparities in resource allocation within the EU. While ethical regulations are generally strong in northern and western countries, Poland and Romania lag behind. This gap may affect the responsible deployment of AI systems and trust in public AI applications. Similarly, public sector adoption rates are sig- nificantly lower in Poland and Romania, potentially due to infrastructural or legislative limitations. The differences underscore the need for targeted sup- port and harmonization efforts by the EU to bridge gaps in AI governance and investment. These results prove hypothesis 6, meaning that all analysed states have AI strategies, but the impact on public administration is different. Enhanced collaboration could foster equitable AI advancements across mem- ber states, reducing digital divides. EU countries such as Germany, Finland, France, Italy, and Spain have AI strategies that include the use of AI in public services. Finland already has a working group to adapt national legislation to the AI Act. Tables 4, 5, and 6 contain comparisons between the main EU coun- tries included in the research. Central European Public Administration Review, Vol. 23, No. 2/2025 84 Armenia Androniceanu, Sofia Elena Colesca Table 5. Comparative approach of the selected EU states from the research parameters perspective Country DESI (2022) ICT graduates % Public sector AI strategies Observations Finland 69.6 high (~>10%) Yes (legal working group) digital champion, robust TH Denmark 69.3 high Yes AI in e- government + governance Netherlands 67.4 high Yes Advanced digital ecosystem Germany 52.9 medium- high Yes AI standardization, active TH Poland 40.6 reduced (~3.5%) Yes Average digital shelter, modest TH Romania 30.6 low (<3%) Yes DIGITIZATION weak, fragile TH Source: authors. Table no. 5 presents a comparative analysis of the main indicators that reflect the main components considered in this research. Based on the results of this research, 3 typologies of approaches to the Triple Helix model in the context of artificial intelligence can be identified. These are: – Leaders: Scandinavia – Finland, Denmark – TH, DESI, solid AI strategies → high AI integration. – Followers: Germany, Netherlands, France – moderate AI integration, but with infrastructure and strategies ready. – Challenged: Poland, Romania – although they have AI strategies, modest digitalization, and innovation ecosystems reduce the impact of AI in admi- nistration. For policymakers, this study highlights the importance of multi-sectoral col- laboration in implementing the EU AI Act; for scholars, it proposes a concep- tual framework linking AI governance to the Triple Helix innovation model. The adoption of artificial intelligence (AI) in public administration presents both significant opportunities and challenges for governance. From a policy perspective, AI can enhance efficiency, streamline decision-making process- es, and improve service delivery. However, successful implementation re- quires robust data governance frameworks, clear regulatory guidelines, and active collaboration between government, industry, and academia, consist- ent with the Triple Helix model. Comparative insights from Estonia, Finland, France, Germany, and Romania highlight diverse approaches: Estonia and Central European Public Administration Review, Vol. 23, No. 2/2025 85 Triple Helix Model and Artificial Intelligence in Public Administration Finland exemplify advanced digital governance infrastructures, while Ger- many and France emphasize stringent data protection and ethical oversight. Romania, still developing its digital governance capabilities, can benefit from adopting best practices from these countries. Ultimately, responsible AI in- tegration in public governance demands a balance between technological innovation, ethical standards, and public accountability to foster trust and equitable service delivery. Table 6. The main research comparative indicators Country DESI (2022) % ICT graduates (2022) Adoption and integration AI (%) AI strategies in public admin. Observation Finland 69.6 ~11 (high) 16 Mature ecosystem Denmark 69.3 ~8 24 Government AI leader Netherlands 67.4 ~7 ~12 Stable, TH collaborations Sweden 65.2 ~6 ~12 Advanced TH, AI e-government Germany 52.9 ~5 ~10 Industrial complex Belgium 50.3 ~3 ~10 Medium TH Spain 60.8 ~5 ~15 Good digital public services Poland 40.5 ~4 ~5 Digital on the rise Romania 30.6 ~2 <5 Low digital Lithuania 52.7 ~5 ~5 Average digital maturity Note: AI adoption values in the public sector are estimated from AI-Watch data on companies. Source: authors. 4.1 Limitations and Further Research The study has several limitations that should be acknowledged. First, the sam- ple is limited to almost ten European Union countries, which, while selected to provide regional and institutional diversity, may reduce the generalizability of the findings across the broader EU or other global contexts. Second, the analysis is based on data from 2018–2023, a period that captures recent de- velopments in AI adoption and digital governance but may not reflect longer- term trends or policy shifts occurring outside this timeframe. Future research could address these limitations by expanding the sample to include additional EU member states or non-European countries, enhancing the comparative perspective and robustness of findings. Longitudinal studies covering longer timeframes could also provide insights into the evolution of Central European Public Administration Review, Vol. 23, No. 2/2025 86 Armenia Androniceanu, Sofia Elena Colesca AI adoption and the impact of governance innovations over time. Additional- ly, future work could explore more granular or sector-specific analyses, exam- ining the effects of AI adoption on particular public administration functions or policy areas. 5 Conclusions This study investigates the relationship between the Triple Helix model (uni- versity–industry–government collaboration) and the assimilation of artificial intelligence (AI) in public administration across selected European Union (EU) member states. The findings highlight distinct cluster patterns among select- ed EU countries, reflecting varying degrees of Triple Helix collaboration, digi- tal maturity, and AI integration in public administration. The study offers valu- able insights into how multi-actor ecosystems can accelerate AI-driven public sector innovation, and it proposes policy recommendations to strengthen cross-sector collaboration for digital governance. This study advances the understanding of how university–industry–government collaboration (Triple Helix) shapes the assimilation of artificial intelligence (AI) within public admin- istration across EU member states. The empirical findings confirm that coun- tries with higher levels of Triple Helix intensity exhibit significantly greater integration of AI applications in public governance processes. The analysis also highlights the mediating effect of digital maturity (measured by DESI) in amplifying the benefits of collaborative innovation ecosystems. The comparative correlation analysis reveals structural disparities across EU countries, with a clear divide between digitally mature states fostering cross- sectoral innovation and lagging countries constrained by weak institutional capacities, limited public-private partnerships, and insufficient ethical govern- ance frameworks. Theoretically, the study contributes to extending the Triple Helix model to the domain of AI-driven public sector innovation, offering a novel perspective on how collaborative ecosystems accelerate digital transformation. Practically, the findings suggest that policymakers should strengthen multi-actor govern- ance mechanisms, integrate ethical AI frameworks into digital strategies, and invest in public sector capabilities to leverage AI for citizen-centric innovation. This study underscores the transformative potential of integrating the Tri- ple Helix Model with Artificial Intelligence to advance innovation, efficiency, and responsiveness in public administration. By aligning the collaborative capacities of academia, industry, and government with data-driven decision- making and intelligent systems, public institutions can transition toward more adaptive and evidence-based governance models. The research con- tributes to the theoretical enrichment of innovation governance and pro- vides practical implications for policymakers aiming to harness AI responsi- bly within the public sector. This paper emphasizes the added value and originality of combining the Tri- ple Helix model with Artificial Intelligence as a framework for reimagining Central European Public Administration Review, Vol. 23, No. 2/2025 87 Triple Helix Model and Artificial Intelligence in Public Administration innovation and governance in public administration. Conceptually, the study advances the field by linking collaborative innovation theory with emerging models of AI governance, providing a fresh perspective on how knowledge co-creation among academia, industry, and government can be strengthened through the use of intelligent technologies. From a practical standpoint, the findings provide actionable insights for poli- cymakers and institutional leaders seeking to foster data-driven, transparent, and citizen-oriented governance. Overall, the research enriches the academic discourse on digital transformation in the public sector and opens new di- rections for empirical investigation into the long-term societal impacts of AI- enabled collaboration. Acknowledgement: The paper has been prepared under the research project named: The impact of artificial intelligence in public administration and algorith- mic governance of human resources in the digital age: opportunities and chal- lenges, 2025, financed by the Bucharest University of Economic Studies. Central European Public Administration Review, Vol. 23, No. 2/2025 88 Armenia Androniceanu, Sofia Elena Colesca References Alaa, A. (2023). Adoption of artificial intelligence and robotics in organisations: a systematic literature review. Int J Bus Technol Manag. 5(3), pp. 342–359. https://doi.org/10.5505/ijbtm.2023.5.3.28 Alon – Barkat, S. and Busuioc, M. (2023). Human–AI Interactions in Public Sector Decision Making: “Automation Bias” and “Selective Adherence” to Algorithmic Advice. Journal of Public Administration Research and Theory, 33(1), pp. 153–169. https://doi.org/10.1093/jopart/muac007 Alon-Barkat, S., and Busuioc, M. (2021). Human–AI Interactions in Public Sector DecisionMaking: Automation Bias and Selective Adherence. Journal of Public Administration Research and Theory, 31(3), pp. 530–547. Androniceanu, A. (2025). The impact of artificial intelligence on human resources processes in public administration. Administratie si Management Public, 44, pp. 75–93. https://doi.org/10.24818/amp/2025.44-05 Androniceanu, A. (2024). Generative artificial intelligence, present and perspectives in public administration. Administratie si Management Public, 43, pp. 105–119. https://doi.org/10.24818/amp/2024.43-06?? Androniceanu, M. (2025). Efficiency and prediction in human resource management using Python modules. Theoretical and Empirical Researches in Urban Management, 20(1), pp. 88–103. Androniceanu, M. (2024). The Alfresco platform, a viable and sustainable strategic option for document management. Management Research and Practice, 16(1), pp. 46–54. Androniceanu, A., and Georgescu, I. (2023). Public administration digitalization and government effectiveness in EU countries. Central European Public Administration Review, 21(1), pp. 7–30. https://doi.org/10.17573/ cepar.2023.1.01 Androniceanu, A., Georgescu, I. and Sabie, O.-M. (2022). The impact of digitalization on public administration, economic development, and well- being in the EU countries. Central European Public Administration Review, 20(1), pp. 7–29. Berman, A., Fine Licht, K. and Carlsson, V. (2024). Trustworthy AI in the public sector: An empirical analysis of a Swedish labor market decision-support system. Technology in Society, 76, p. 102471. https://doi.org/10.1016/j. techsoc.2024.102471. Carayannis, E.G. and Campbell, D.F.J. (2009). ’Mode 3’ and ’Quadruple Helix’: Toward a 21st Century Fractal Innovation Ecosystem. International Journal of Technology Management, 46(3/4), pp. 201–234. European Commission. (2023). Europe’s digital future: 2023 Report on the state of the Digital Decade. At , accessed 3 July 2025. De Alwis, A.C., Andrlić, B. and Šostar, M (2022). The influence of E-HRM on modernizing the role of HRM context. Economies. 10(8), pp. 181–195. https:// doi.org/10.3390/economies10080181.11. Dvorsky, J. (2025). Impact of Artificial Intelligence on Enterprise Risk Mana- gement. A case study from the Slovak SME Segment. Journal of Business Sectors, 3(1), pp. 96–103. Etzkowitz, H. (2003a). Innovation in innovation: the Triple Helix of university- industry-government relation. Social Science Information, 42(3), pp. 293–338. Central European Public Administration Review, Vol. 23, No. 2/2025 89 Triple Helix Model and Artificial Intelligence in Public Administration Etzkowitz, H. (2008), Triple Helix Innovation: Industry, University, and Government in Action, London and New York: Routledge. Etzkowitz, H. and Carvalho de Mello, J. M. (2004). The rise of a Triple Helix culture: Innovation in Brazilian economic and social development. Inter- national Journal of Technology Management and Sustainable Development, 2-3, pp. 159–171. Etzkowitz, H. and Klofsten, M. (2005). The innovating region: toward a theory of knowledge-based regional development. RandD Management, 35(3), pp. 43–255. Etzkowitz, H. (2003b). Innovation in innovation: the Triple Helix of university- industry-government relation’. Social Science Information, 42(3), pp. 293–338. Etzkowitz, H. and Leydesdorff, L. (2001). The Transformation of University- Industry-Government Relations. Electronic Journal of Sociology (online), 5(4), pp. 1–32. At , accessed 4 June 2025. Etzkowitz, H. and Leydesdorf, L. (1999). The future location of research and technology transfer. Journal of Technology Transfer, 24 (2-3), pp. 111–123. Etzkowitz, H. and Leydesdorff, L. (1995a). The Triple Helix: university -industry – government relations. A laboratory for knowledge based economic development. EASST Review. European Society for the Study of Science and Technology, 14(1), pp. 18–36. Etzkowitz, H. and Leydesdorff, L. (1995b). The Triple Helix: University–Industry– Government Relations. Journal of Knowledge-based Innovation in China, 1(2), pp. 107–123. European Union (2022a). Digital Economy and Society Index (DESI) 2022. Retrieved from here: https://digital-strategy.ec.europa.eu/en/library/digital- economy-and-society-index-desi-2022. Accessed on 17 October 2025. European Commission (2022b). AI Watch. National strategies on Artificial Intelligence: A European perspective. 2022 edition. At , accessed 10 June 2025. Garg, S. et al. (2022). A review of machine learning applications in human resource management. Int J Product Perform Manag., 71(5), pp. 590–610. https://doi.org/10.1108/IJPPM-08-2020-0427. Gill, S.S. et al. (2024). Transformative effects of ChatGPT on modern education: Emerging Era of AI Chatbots. Internet of Things and Cyber-Physical Systems, 4, pp. 19–23. https://doi.org/10.1016/j.iotcps.2023.06.002 Grilli, L. and Pedota, M. (2024). Creativity and artificial intelligence: A multilevel perspective. Creativity and Innovation Management, 33(2), pp. 234–247. https://doi. org/10.1111/caim.12580. Joint Research Centre. (2024). Competences and Governance Practices for Artificial Intelligence in the Public Sector. Technical Report. At , accessed 1 July 2025. Kern, A. et al. (2021). Fairness in Algorithmic Profiling: A German Case Study. Chalmers University Research Report, 1–33. At , accessed 1 July 2025. Kruger, S. and Steyn, A.A. (2025). Navigating the fourth industrial revolution: a systematic review of technology adoption model trends. Journal of Science and Technology Policy Management, 16(10), pp. 24–56. https://doi. org/10.1108/JSTPM-11-2022-0188. Central European Public Administration Review, Vol. 23, No. 2/2025 90 Armenia Androniceanu, Sofia Elena Colesca Leydesdorff, L. (2009). The knowledge-based economy and the triple helix model. Annual Review of Information Science and Technology, 44(1), pp. 365–417. https://doi.org/10.1002/aris.2010.1440440116. Accessed on 26 June 2025. Lorincova, S. et al. (2024). Identifying Corporate Culture Using the Organizational Culture Assessment Instrument. Journal of Business Sectors, 2 (1), pp. 11–20. Malin, C. et al. (2023). In the AI of the beholder—a qualitative study of HR professionals’ beliefs about AI-based chatbots and decision support in candidate pre-selection. Administrative Science. 13(11), p. 231. https://doi. org/10.3390/admsci13110231. Makridakis, S. (2017). The forthcoming Artificial Intelligence (AI) revolution: its impact on society and firms. Futures, 90, pp. 46–60. https:// doi. org/10.1016/j.futures.2017.03.006. Meshram, R. (2023). The role of artificial intelligence (AI) in recruitment and selection of employees in the organisation. Russian Law Rev, 11(9), pp. 322– 333. Michalec, G., Hargitai, D. M., and Bencsik, A. (2024). Organizational trust as a success factor. Journal of Business Sectors, 2(1), pp. 61–67. Mwita, K. M. and Kitole, F. A. (2025). Potential benefits and challenges of artificial intelligence in human resource management in public institutions. Discover Global Society, 3(35), pp. 1-19. https://doi.org/10.1007/s44282-025- 00175-8. At , accessed 7 August 2025. Neumann. O., Guirguis, K. and Steiner, R. (2024). Exploring artificial intelligence adoption in public organizations: a comparative case study. Public Management Review, 26(1), pp. 114–141. https://doi.org/10.1080/14719037 .2022.2048685. Nosratabadi S. et al. (2019). Sustainable Business Models: A Review. Sustainability, 11(6), p. 1663. https://doi.org/10.3390/su11061663 Nyathani, R. (2023). AI in performance management: redefining performance appraisals in the digital age. Journal of Artificial Intelligence and Cloud Computing, 2(4), pp. 1–5. https://doi.org/10.4736/JAICC/2023(2)134. OECD. (2022). Artificial Intelligence. At , accessed 14 October 2025. Oxford Insights (2023). Government AI Readiness Index 2022. At , accessed on 14 October 2025. Panagiotopoulos, P., Bowen, F. and Protopsaltis, S. (2019). Technological Innovation and Public Value: Dissemination and Challenges. London: Edward Elgar Publishing. Parent-Rocheleau, X. and Parker, S.K. (2022). Algorithms as work designers: How algorithmic management influences the design of jobs. Hum Resour Manag Rev, 32(3), p. 100838. https://doi.org/10.1016/j.hrmr.2021.100838. Rahman M. and Aydin E. (2020). Benefits, barriers and risks – the role of technology in e-HRM implementations in public sector organisations: evidence from Bangladesh. Int J Human Res Dev Manag, 20(3), pp. 252–268. Ranga, M., and Etzkowitz, H. (2013). Triple Helix systems: an analytical framework for innovation policy and practice in the Knowledge Society. Industry and higher education, 27(4), pp. 237–262. Central European Public Administration Review, Vol. 23, No. 2/2025 91 Triple Helix Model and Artificial Intelligence in Public Administration Reis, J., Santo, P.E., and Melão, N. (2019). Artificial Intelligence in Government Services: A Systematic Literature Review. In Rocha, Á. et al., eds., New Knowledge in Information Systems and Technologies. Advances in Intelligent Systems and Computing. Cham: Springer, pp. 241–252. https://doi. org/10.1007/978-3-030-16181-1_23 Riccio, P. et al. (2022). Algorithmic Censorship of Art: A Proposed Research Agenda. At , accessed 12 August 2025 Rodgers, W. et al. (2024). An artificial intelligence algorithmic approach to ethical decision-making in human resource management processes. Hum Resour Manag Rev, 33(1), p. 100925. https://doi.org/10.1016/j.hrmr.2022.100925. Stachová K. et al. (2024). The impact of E-HRM tools on employee engagement. Administrative Science, 14(11), p. 303. https://doi. org/10.3390/ admsci14110303. Straub, V. J. et al. (2022). Artificial Intelligence in Government: The AIGOV Framework. Journal of Public Administration, 24(2), pp. 145–168. Strohmeier S. (2020). Digital human resource management: a conceptual clarification. German J Human Res Manag: Zeitschrift Für Personalforschung, 34(3), pp. 345–365. https://doi.org/10.1177/2397002220921131. Tangi, L. et al. (2023). Artificial Intelligence for Interoperability in the European Public Sector – an Exploratory study. At , accessed 1 July, 2025. Votto, A.M. et al. (2021). Artificial intelligence in tactical human resource management: a systematic literature review. International Journal of Information Management and Data Insights, 1(2), p. 100047. https://doi. org/10.1016/j.jjimei.2021.100047. Zuiderwijk, A., Chen, Y.-C. and Salem, F. (2021). The Impact of AI on Public Sector Transformation: A Systematic Review. Government Information Quarterly, 38(4), p. 101570. Wirtz, B. W., Weyerer, J. C. and Geyer, C. (2018). Artificial Intelligence and the Public Sector—Applications and Challenges. International Journal of Public Administration, 42(7), pp. 596–615. https://doi.org/10.1080/01900692.2018. 1498103