Norwegian version of this page

Pouria Akbarighatar

Assistant Professor Department of Information Systems

Email: pouria.akbarighatar@uia.no
Phone: +47 38 14 26 26
Office: H1018 (Universitetsveien 25, 4630 Kristiansand, Norway)

Press photo Download business card

Background and experience

University of Agder, Ph.D. in Information Systems (2022 – 2026)

Data Scientists, Telecommunication Industry (2016 – 2022)

University of Tehran, MA in IT Management (2015 – 2017)

Bu-Ali Sina University, BS in Industrial Engineering (2009 – 2013)

Areas of responsibility

I am interested in helping organizations create sustainable and reliable value from technology. My research focuses on two key themes: a) the impact of artificial intelligence on society, particularly questions of responsibility in algorithmic decision-making, the inscrutability of complex AI models, and potential biases arising from data; b) the use and value of data for decision-making, as well as how data science can help organizations better manage the risks associated with data-driven decisions.

Teaching

Research

Selected publications

Akbarighatar, P., Pappas, I. and Vassilakopoulou, P., 2023. A sociotechnical perspective for responsible AI maturity models: Findings from a mixed-method literature review. International Journal of Information Management Data Insights , 3 (2), p.100193. ISSN: 2667-0968. doi: https://doi.org/10.1016/j.jjimei.2023.100193
Jafarzadeh, Hamed; Heidary-Dahooie, Jalil; Akbarighatar, Pouria (2022). A project prioritization approach considering uncertainty, reliability, criteria prioritization, and robustness. Decision Support Systems . ISSN: 0167-9236. 156 doi: 10.1016/j.dss.2022.113731 .
Akbarighatar, Pouria; Jafarzadeh, Hamed (2018). A methodology for project portfolio selection under criteria prioritisation, uncertainty and projects interdependency–combination of fuzzy QFD and DEA. Expert Systems With Applications . ISSN: 0957-4174. doi: 10.1016/j.eswa.2018.05.028 .
Akbarighatar, P., 2024. Operationalizing responsible AI principles through responsible AI capabilities. AI and Ethics , pp.1-15. doi: https://doi.org/10.1007/s43681-024-00524-4

Research groups

Akbarighatar, Pouria; Pappas, Ilias & Vassilakopoulou, Polyxeni (2025). Responsible AI in Action: Assessing Variations in Perspectives Between C-Level Managers and AI Developers. In Brooks, Laurence; Li, Honglei & Papagiannidis, Savvas (Ed.), Proceedings of the 30th UK Academy for Information Systems (UKAIS) International Conference. Association for Information Systems (AIS). ISSN 9781739087524. p. 242–261. doi: https:/www.ukais.org/wp-content/uploads/2025/04/UKAIS-2025-full-proceedings-ver1.pdf. Full text in Research Archive Show summary
Organizations worldwide embrace Artificial Intelligence (AI) initiatives across diverse fields, from healthcare to transportation. While AI may offer significant benefits, there are also concerns related to issues such as bias, privacy, transparency, reliability, and accountability. Responsible AI (RAI) frameworks and principles have emerged to address these issues. However, operationalizing RAI remains a challenge. In this paper, we report findings from a survey of C-level managers and AI developers from Europe and North America. We explore the perceptions of these different practitioner groups and identify areas of consensus and divergence. The results suggest that organizations may face challenges in effectively actualize the RAI, due to diverging perspectives. By highlighting these challenges and insights, this study contributes a nuanced understanding of the dynamics influencing real-world RAI and can contribute to better intraorganizational communication and collaboration.
Akbarighatar, Pouria; Rinta-Kahila, Tapani & Someh, Ida (2025). When Welfare Goes Digital: Lessons from the Dutch SyRI Risk Indicator in Public Sector. Academy of Management Proceedings. 2025(1). doi: 10.5465/amproc.2025.292bp. Full text in Research Archive Show summary
Algorithmic decision-making (ADM) systems, designed to enhance efficiency and fairness in public service delivery, are often portrayed as tools to improve social protection programs. However, this optimistic view is challenged by evidence showing that these systems can introduce bias, lack transparency, and produce inequitable outcomes for vulnerable populations. This paper examines ADM systems through the lens of data justice, which focuses on fairness in how individuals are represented, treated, and impacted by digital systems. Using a data justice framework, we analyze a case study of the Dutch System Risk Indication (SyRI), focusing on the legal, informational, and design-related forms of injustice that emerge in its application. Our findings highlight systemic inequities inherent in such systems, including exclusion errors that deny rightful benefits and inclusion errors that misallocate resources. By highlighting these issues, this paper contributes to the ongoing discourse on the ethical implications of ADM systems, offering insights into how data injustices can be identified and mitigated in public sector applications.
Akbarighatar, Pouria; Pappas, Ilias; Purao, Sandeep Ramakant & Vassilakopoulou, Polyxeni (2025). Enacting Responsible AI: A Configurational Analysis of AI Principles in Practice. Information Systems Frontiers. ISSN 1387-3326. doi: 10.1007/s10796-025-10618-x. Full text in Research Archive Show summary
Responsible AI (RAI) entails developing, using, and governing AI in a human-centred way to ensure its trustworthiness and alignment with human values. Organizations attempt to enact RAI through guiding principles that aim to minimize threats such as bias and privacy violations and enhance outcomes such as transparency and fairness. In the empirical study we report in this paper, we share insights about how AI experts view RAI principles, including relationships among them, these principles, and the enactment of RAI principles. Employing a sequential mixed-methods approach, we analyse survey data from 135 AI experts from Europe and America using fuzzy-set Qualitative Comparative Analysis (fsQCA), complemented by follow-up interviews. Our analysis identifies five equifinal combinations of RAI principles associated with high overall RAI enactment. The findings reveal the crucial role of accountability and safety-reliability principles and highlight the contextual variability of other principles. This research contributes a nuanced understanding of RAI enactment, moving beyond individual principles to demonstrate their complex interplay and their contextual dependencies in achieving responsible AI practices.
Akbarighatar, Pouria (2024). Operationalizing responsible AI principles through responsible AI capabilities. AI and Ethics. ISSN 2730-5953. p. 1–15. doi: 10.1007/s43681-024-00524-4. Full text in Research Archive
Akbarighatar, Pouria; Pappas, Ilias; Vassilakopoulou, Polyxeni & Purao, Sandeep Ramakant (2024). Responsible AI Principles: Findings from an Empirical Study on Practitioners' Perceptions. In Laurence, Brooks; Angelopoulos, Spyros; Pan, Yu-Chun & Zamani, Efpraxia D. (Ed.), Proceedings of the 29th UK Academy for Information Systems (UKAIS) International Conference. Association for Information Systems (AIS). ISSN 9781739087517. Full text in Research Archive
Abedin, Ehsan; Mendoza, Antonette; Akbarighatar, Pouria & Karunasekera, Shanika (2024). Predicting Credibility of Online Reviews: An Integrated Approach. IEEE Access. 12, p. 49050–49061. doi: 10.1109/ACCESS.2024.3383846. Full text in Research Archive
Akbarighatar, Pouria; Pappas, Ilias & Vassilakopoulou, Polyxeni (2023). Practices for Responsible AI: Findings from Interviews with Experts. In Midha, Vishal & Animesh, Animesh (Ed.), The 29th Americas Conference on Information Systems (AMCIS). Association for Information Systems (AIS). ISSN 9781958200056. Full text in Research Archive
Akbarighatar, Pouria; Pappas, Ilias & Vasilakopoulou, Polyxeni (2023). A sociotechnical perspective for responsible AI maturity models: Findings from a mixed-method literature review. International Journal of Information Management Data Insights. 3(2). doi: 10.1016/j.jjimei.2023.100193. Full text in Research Archive Show summary
As artificial intelligence (AI) is increasingly used in various industries, it becomes crucial for organizations to enhance their capabilities and maturity in adopting AI responsibly. This paper employs a mixed-method approach that combines topic modeling with manual content analysis to provide a comprehensive review of the literature on AI maturity and readiness. The review encompasses an extensive corpus of 1451 papers, identifying the main themes and topics within this body of literature. Based on these findings, a subset of papers was selected and further analyzed to identify AI capabilities utilizing a sociotechnical lens. This further analysis led to the identification of foundational and responsible AI (RAI) capabilities. These capabilities have been integrated in a sociotechnical framework of capabilities for AI maturity models providing valuable insights for organizations and AI service providers and a basis for further research.
Akbarighatar, Pouria; Pappas, Ilias & Vasilakopoulou, Polyxeni (2023). JUSTICE AS FAIRNESS: A HIERARCHICAL FRAMEWORK OF RESPONSIBLE AI PRINCIPLES, Proceedings of the European Conference on Information Systems ECIS 2023. Association for Information Systems (AIS). ISSN 9781958200087. Full text in Research Archive
Jafarzadeh, Hamed; Heidary-Dahooie, Jalil & Akbarighatar, Pouria (2022). A project prioritization approach considering uncertainty, reliability, criteria prioritization, and robustness. Decision Support Systems. ISSN 0167-9236. 156. doi: 10.1016/j.dss.2022.113731. Full text in Research Archive Show summary
Given the limitation of resources in organizations (time, money, staff, material, etc.), deciding which projects should be given priority – among many competing projects – is a constant and continuous challenge for decision makers. As a result, a proliferation of methods and solutions for project prioritization has been developed, each with its own strengths and limitations. In this paper, we propose a hybrid approach for prioritizing projects taking into consideration four pervasive challenges in decision making: 1) uncertainty (imprecision and vagueness) of human decisions, 2) reliability (extent of sureness) of decision makers, 3) systematic identification of selection criteria, and 4) robustness of decision making (reasonable tolerance against insignificant changes to decision makers’ evaluations during the process). In doing so, we propose integrating Z-numbers theory with Quality Function Development (ZQFD), four Multi-Criteria Decision-Making (Z-MCDM) methods, ensemble ranking aggregation, and sensitivity analysis. The proposed method is then applied in a real-world organization with twenty IT projects to illustrate how our approach might be used in practice. Furthermore, we elaborate on the trustworthiness of the proposed method in light of 16 criteria available in the literature of structured decision making.
Akbarighatar, Pouria (2022). MATURITY AND READINESS MODELS FOR RESPONSIBLE ARTIFICIAL INTELLIGENCE (RAI): A SYSTEMATIC LITERATURE REVIEW, The 14th Mediterranean Conference on Information Systems (MCIS 2022). Acting in the digital society: moving towards a sustainable future. Association for Information Systems (AIS). ISSN 9788887173055. p. 1–17. Full text in Research Archive Show summary
The development and deployment of Artificial intelligence (AI) in organizations is of growing interest to the Information systems (IS) discipline. This can be approached from a sociotechnical perspective contributing to managing the unintended outcomes of AI while extending AI use boundaries. This paper presents the findings from a systematic literature review on organizational maturity and readiness for AI development and development. A key result is that extant research has lost sight of AI systems' humanistic and ethical aspects, and principles related to responsible AI are not sufficiently defined. This is a hurdle because principles for responsible AI are fundamental for AI development and deployment ensuring long-term benefits. Drawing from the literature review findings, we provide a conceptual maturity model with two main dimensions (responsible and instrumental), twelve conditions and thirty factors. The maturity and readiness factors for responsible AI are deduced from synthesizing 35 articles in related literature. Specifically, the paper identified six capabilities for responsible AI: AI model, Cooperative AI, ethical awareness, laws & regulations, data governance and continuous improvement; six instrumental capabilities were also identified: strategic alignment, technology, culture, data management, financial and human resource management.
Akbarighatar, Pouria (2020). Prioritising IT Projects: Combination of Fuzzy QFD and ARAS to Address Criteria Multiplicity Challenge, Proceedings of the 23rd Pacific Asia Conference on Information Systems (PACIS). Association of Information Systems. ISSN 9781733632539. Full text in Research Archive
Akbarighatar, Pouria (2020). Prioritising IT projects: A multi-method approach, Proceedings of the 23rd Pacific Asia Conference on Information Systems (PACIS). Association of Information Systems. ISSN 9781733632539. Full text in Research Archive
Akbarighatar, Pouria & Jafarzadeh, Hamed (2018). A methodology for project portfolio selection under criteria prioritisation, uncertainty and projects interdependency–combination of fuzzy QFD and DEA. Expert Systems With Applications. ISSN 0957-4174. doi: 10.1016/j.eswa.2018.05.028. Full text in Research Archive

View all works in NVA

Sarabadani, Jalal & Akbarighatar, Pouria (2025). AIS Electronic Library (AISeL) - AMCIS 2025 Proceedings: Unpacking Students’ Irresponsible Use of Generative AI in Higher Education: A Review of Antecedents and Consequences. AMCIS 2025. Full text in Research Archive Show summary
The widespread use of Generative AI in higher education has raised concerns about its irresponsible use, as its misuse threatens learning outcomes and professional readiness. This paper aims to synthesize existing research on students’ irresponsible use of GAI in higher education. Through a literature review, we identified five primary forms of irresponsible use: (1) plagiarism, (2) misinformation and hallucination, (3) over-reliance on GAI, (4) transparency and accountability issues, and (5) privacy risks and security concerns. Our findings reveal that these behaviors are driven by individual, institutional and technology factors with negative consequences on students’ cognitive development and future employability. Our study underscores the urgent need for institutions to establish GAI policies and literacy, as a few actions to promote responsible use of GAI. We contribute to the ongoing discussions on academic integrity and provide a foundation for future research on mitigating the risks associated with GAI misuse in higher education.
Akbarighatar, Pouria; Pappas, Ilias & Vassilakopoulou, Polyxeni (2024). Are AI practitioners ready for AI Fairness? The need for Institutional Work for Early Prioritization of Fairness in AI Practices. Full text in Research Archive
Akbarighatar, Pouria; Pappas, Ilias & Vassilakopoulou, Polyxeni (2023). Justice as Fairness: a hierarchical framework of Responsible AI principles. Full text in Research Archive
Akbarighatar, Pouria (2026). Enacting Responsible AI in Organizations: A Sociotechnical Perspective. Universitetet i Agder. ISSN 9788284273037. Full text in Research Archive Show summary
Forskningskontekst – Organisasjoner i både offentlig og privat sektor tar i økende grad i bruk kunstig intelligens (KI) til beslutningsstøtte, automatisert beslutningstaking og dataanalyse for både beskrivende og prediktive formål. Disse applikasjonene har ofte betydelig innvirkning på enkeltpersoner, eller brukes i industrielle sammenhenger med brede driftsmessige konsekvenser. Prinsipper for ansvarlig kunstig intelligens (RAI) kan fungere som verktøy for å redusere utilsiktede konsekvenser av å ta i bruk slike systemer i organisasjoner. Motivasjon og formål – RAI har i økende grad blitt en høy prioritet for organisasjoner, spesielt med tanke på økende regulatoriske krav, slik som EU AI Act og økende bevissthet rundt etiske hensyn. I denne sammenhengen er den dominerende tilnærmingen til RAI begrenset av en antakelse om å «fylle gapet» mellom prinsipper og praksis. Dette skiller abstrakte prinsipper fra praktisk implementering og fokuserer for snevert på verktøy og rammeverk, samtidig som sosiotekniske kompleksiteter neglisjeres. Videre legger eksisterende forskning større vekt på overordnede prinsipper enn på hvordan disse kan gjøres handlingsrettede for organisasjonsaktører som utviklere og ledere. I en tid preget av raske endringer bør oversettelsen av RAI-prinsipper til praksis ikke ses som en engangsinnsats, men som en kontinuerlig prosess for å fremme ansvarlighet i organisasjoner. Formålet med denne avhandlingen er å undersøke implementeringen av RAI i organisasjoner fra et sosioteknisk perspektiv, og å bringe frem innsikt som kan være nyttig for både forskning og praksis. Forskningsdesign – Jeg har brukt en flermetodetilnærming for å svare på forskningsspørsmålene og konseptualisere RAI-implementering. En abduktiv resonnementstilnærming har blitt brukt gjennom hele avhandlingen, noe som muliggjør en iterativ veksling mellom empiriske observasjoner og teoretisk informert tolkning. For å gjøre dette, har jeg først gjennomgått litteraturen om RAI-modenhets- og beredskapsmodeller, for deretter å gjennomføre den første empiriske studien (Studie 1). Denne besto av to runder med kvalitative intervjuer med eksperter som har relevant erfaring med implementering av RAI fra USA og Europa, og baserer seg på 20 intervjuer og 348 sider med offentlige rapporter for implementering av RAI. Studie 1 ga innsikt i kompleksiteten i å oversette RAIprinsipper til praksis. Studie 2 ble utført med en fuzzy-set kvalitativ komparativ analyse (fsQCA) av svar fra 135 ekspertrespondenter. Målet var å undersøke hvordan RAI-prinsipper vurderes samlet i praksis og å utforske deres konfigurasjonsmessige innvirkning, i stedet for å analysere hvert prinsipp isolert. Et spørreundersøkelsesinstrument ble utviklet og validert for pålitelighet for å støtte denne fasen. Til slutt ble en tredje studie (studie 3) kontaktet, hvor datasettet med svar fra eksperter ble videre analysert ved å fokusere på deltakere fra to grupper: ledere og AI-utviklere. En ikke-parametrisk analyse ble utført for å undersøke om det var signifikante forskjeller i oppfatninger av RAI-prinsipper og deres implementering på tvers av ulike organisasjonsroller. Funn – Denne forskningen tilbyr en konseptualisering av RAI-kapasitet som gjør det mulig for organisasjoner å bevege seg mot implementering av RAI gjennom to distinkte utviklingsveier. De empiriske funnene demonstrerer også konseptet med ekvivalens i implementeringen av RAI. Dette innebærer at det er mulig å oppnå samme resultat (høye nivåer av RAI-implementering) gjennom ulike veier. Gjennom anvendelsen av fsQCA identifiserer studien fem ekvivalente konfigurasjoner av RAI-prinsipper som kan føre til RAI-implementering. Disse understreker den kontekstuelle og sosiotekniske naturen ved RAI-implementering: relevansen og prioriteringen av prinsipper kan variere avhengig av domenet der KI tas i bruk. Teoretiske implikasjoner – Denne avhandlingen gir betydelige bidrag til forskning. For det første, ved å bruke en konfigurasjonsbasert tilnærming ved bruk av fsQCA, fanger den opp kompleksiteten og gjensidige avhengigheter mellom ulike forhold som kreves for RAI-implementering. Gjennom denne konfigurasjonsbaserte linsen identifiserer studien ikke bare flere veier som fører til ansvarlige AI-utfall, men bruker også disse veiene som grunnlag for å teoretisere prosessen med RAI-implementering. Denne tilnærmingen går utover lineære årsak-virkning-forhold og tilbyr ny teoretisk innsikt i de mange veiene som organisasjoner implementerer RAI gjennom. For det andre foreslår forskningen et konseptuelt rammeverk ved å integrere innsikt fra funnene fra den konfigurasjonsbaserte tilnærmingen og RBV. RAI-funksjoner, som en muliggjører, spiller en avgjørende rolle i å la organisasjoner ta i bruk og opprettholde flere konfigurasjoner av RAI-prinsipper, og dermed fremme RAI-implementering. Denne prosessen er iboende sosioteknisk, og krever at både AI-utviklere og ledere aktivt forstår prinsippene og omsetter dem til praksis. Dette viser at RAI-prinsipper og -praksis ikke er separate, men sammenvevd og gjensidig innflytelsesrike. Til slutt gir dette rammeverket et generaliserbart, men likevel tilpasningsdyktig perspektiv for å studere RAI-implementering. Praktiske implikasjoner – Basert på det foreslåtte rammeverket for RAI-aspekter og de empiriske funnene som presenteres på tvers av artiklene, er det utviklet åtte praktiske anbefalinger for å veilede organisasjoner i å operasjonalisere ansvarlig AI. Innsikten som genereres er praktisk verdifull for både ledere og AI-eksperter, ettersom den spiller en viktig rolle i å gi et grunnlag for RAI-operasjonalisering. Denne kunnskapen kan hjelpe organisasjoner med å begynne å tenke kritisk om de essensielle aspektene som trengs for å operasjonalisere RAI i hverdagen og å engasjere seg dypere i de spesifikke ferdighetene som kreves for å oppfylle prinsippene som er skissert i eksisterende retningslinjer. Originalitet/verdi – Denne avhandlingen tilbyr en ny og integrerende konseptualisering av RAI-implementering, og fremmer forståelsen av hvordan organisasjoner jobber med abstrakte etiske prinsipper og implementerer dem i situasjonsbestemt praksis. Den rekonseptualiserer RAI-implementering ikke som en statisk samsvarsøvelse, men som å sette ut i livet en utviklende, organisatorisk evne (RAI-evne) som implementeres kontekstuelt. Originaliteten til dette arbeidet ligger i å introdusere begrepet RAI-implementering, som artikuleres gjennom de tre sammenhengende aspektene oversettelse, tilpasning og bærekraft, og i å demonstrere, gjennom både empiriske bevis og teoretisk syntese, at RAI oppstår fra flere, likeverdige veier snarere enn en enkelt universell modell. Denne avhandlingen bidrar dermed til teori ved å kombinere normative retningslinjer med innsikt fra praksis i den virkelige verden, og til praksis ved å tilby handlingsrettet veiledning for organisasjoner som ønsker å være ansvarlige i sine KI-initiativer.

View all works in NVA

Published Apr. 16, 2024 11:32 AM - Last modified Feb. 27, 2026 2:46 PM