artificial intelligence
THE DEPENDENCE OF ARTIFICIAL INTELLIGENCE ON HUMAN INTELLIGENCE: AN ANALYSIS OF THE DIELLA CASE AS THE WORLD’S FIRST AI MINISTER AND THE CORRUPTION SCANDAL IN ALBANIA By MYR Agung Sidayu- Yayasan pendidikan Indonesia –...
artificial intelligence
THE DEPENDENCE OF ARTIFICIAL INTELLIGENCE ON HUMAN INTELLIGENCE: AN ANALYSIS OF THE DIELLA CASE AS THE WORLD’S FIRST AI MINISTER AND THE CORRUPTION SCANDAL IN ALBANIA
By
MYR Agung Sidayu- Yayasan pendidikan Indonesia – Special Consultative Status with ECOSOC – United Nations since 2013
#UnitedNations- #undesa- #ECOSOC – #UNESCO – #kemendikbudri – #kemenag
Important Note
The case of Diella, an artificial intelligence (AI) system appointed as Minister of State for Artificial Intelligence in Albania in September 2025, provides empirical evidence that artificial intelligence is highly dependent on human intelligence. A project promoted as a radical solution to eradicate corruption in public procurement became entangled in a corruption scandal at its developing agency (AKSHI) and a lawsuit concerning personal data violations filed by actress Anila Bisha. Through an analysis of this case, this article affirms that the effectiveness and legitimacy of AI are determined by algorithmic transparency, institutional accountability, and ethical human oversight. Without these elements, AI risks reflecting and amplifying the weaknesses of the humans who create it.
Divine Principles
QS. An-Naḥl [16]: 78
وَاللَّهُ أَخْرَجَكُم مِّن بُطُونِ أُمَّهَاتِكُمْ لَا تَعْلَمُونَ شَيْئًا وَجَعَلَ لَكُمُ السَّمْعَ وَالْأَبْصَارَ وَالْأَفْئِدَةَ ۙ لَعَلَّكُمْ تَشْكُرُونَ
“And Allah brought you out of your mothers’ wombs knowing nothing, and He gave you hearing, sight, and hearts so that you might give thanks.”
QS. Al-‘Ankabūt [29]: 43
وَتِلْكَ الْأَمْثَالُ نَضْرِبُهَا لِلنَّاسِ ۖ وَمَا يَعْقِلُهَا إِلَّا الْعَالِمُونَ
“And these examples We present to the people, but none will understand them except those of knowledge.”
1. Introduction
Yesterday I wrote about AI and education as a consequence that must be accepted with positive thinking. The article entitled “The Integration of AI and ICT in Education: A Deweyan Perspective, Policy Analysis, and Recent Developments Toward Inclusive and Transformative Education” constitutes a conclusion drawn from one chapter of my dissertation, “A Historical and Philosophical Analysis of the Modern Philippine Educational System: Legal Foundations and Policy Reforms.”
I was very pleased that my colleague Prof. Dr. M. Tolhah (UIN Sunan Ampel Surabaya) offered a highly scholarly comment, stating that no matter how sophisticated and meaningful artificial intelligence may be, it still requires the intervention of human intelligence. This observation is entirely correct, and it has now been empirically demonstrated by the scandal that occurred in Albania.
The Diella case—an AI system appointed as “Minister of State for Artificial Intelligence” in September 2025—has become a living laboratory testing that thesis. Diella was promoted as a solution to eliminate corruption in public procurement, a sector long plagued by scandals in Albania. Yet within less than a year, the project faced allegations of large-scale corruption at the National Agency for Information Society (AKSHI) as well as legal action concerning personal data violations.
This article aims to analyze the Diella case in order to demonstrate that AI is not autonomous. Its effectiveness, legitimacy, and integrity depend heavily on human intelligence, manifested through algorithmic transparency, accountability, and ethical oversight. The analysis is relevant to the broader discourse on the integration of AI in governance and education.
2. Theoretical Framework
The literature on AI ethics emphasizes the importance of human-in-the-loop approaches and algorithmic accountability (Floridi et al., 2018; Mittelstadt et al., 2016). AI is a human artifact; design decisions, training data, and parameters are entirely determined by humans. Without algorithmic transparency (explainability and auditability), AI becomes a “black box” that can amplify existing biases or problematic practices.
John Dewey’s perspective on technology as a tool for reflective experience and democracy underscores that innovation must be grounded in critical human deliberation. In the context of governance, AI is effective only when controlled by robust ethical frameworks and institutions.
This study employs a qualitative case-study approach with document analysis. Primary and secondary data were obtained from international media reports (Reuters, OCCRP, BBC, and Politico), official documents of the Albanian government, and anti-corruption investigation reports. The analysis is descriptive-analytical in nature, aiming to reveal the dynamics of Diella’s development, implementation, and controversies, as well as their implications for AI’s dependence on human intelligence.
4. Results and Discussion
4.1 Background and Development of Diella
Albania has long faced systemic corruption challenges, particularly in the public procurement sector. This sector has not only been a major source of state budget leakage but also a critical obstacle in the European Union accession process targeted by the Tirana government around 2030. Transparency International’s Corruption Perceptions Index has consistently ranked Albania in a problematic position, where practices of nepotism, bribery, and conflicts of interest in government tenders have formed a recurring pattern for decades.
As part of its digitalization and governance reform agenda, on 19 January 2025 the National Agency for Information Society (AKSHI) officially launched Diella (meaning “sun” in Albanian) as a virtual assistant on the e-Albania platform. The system was developed through collaboration between AKSHI and Microsoft, utilizing OpenAI’s large language model running on Azure infrastructure. In its initial phase (Diella 1.0), the system functioned as a text-based chatbot assisting citizens in navigating digital public services. It was later upgraded to Diella 2.0 with voice interaction capabilities and an animated avatar.
Official data and media reports indicate that by mid-2025, Diella had facilitated more than one million interactions and applications and had assisted in the issuance of more than 36,000 digital documents. Diella’s avatar is depicted as a young woman dressed in traditional attire from the Zadrima region (northern Albania), with her face and voice taken from film and theater actress Anila Bisha.
The culmination of the transformation occurred on 11 September 2025. Four months after Prime Minister Edi Rama’s fourth electoral victory in May 2025, Diella was officially appointed “Minister of State for Artificial Intelligence”—making it the world’s first AI system to hold a cabinet-level position. Its primary mandate was to oversee all public tender processes to ensure they were “100 percent free of corruption.” Rama stated that decisions regarding tender requirements, eligibility criteria, price ceilings, and winners would gradually be transferred from human ministries to the AI system in order to guarantee objectivity, speed, and transparency.
This step was positioned as a global breakthrough and a response to European Union demands regarding anti-corruption measures. From the outset, however, fundamental questions arose: to what extent can a system developed within a state institution truly remain independent of political influence and the interests underlying its development?
4.2 The Anila Bisha Lawsuit: Personal Data Violations and Failures of Ethical Governance
In February 2026, the Diella controversy entered a deeper legal dimension. Anila Bisha, the actress whose face and voice were used for the Diella avatar, filed a lawsuit at the Tirana Administrative Court. The defendants included the Council of Ministers of the Republic of Albania, AKSHI, the private company Apel shpk (involved in avatar production), and Prime Minister Edi Rama personally.
The core of Bisha’s lawsuit concerns contract and personal data violations. According to her, in late 2024 she signed an agreement authorizing the use of her face and voice for only one year (until 31 December 2025) and solely for the virtual assistant function on the e-Albania portal. When Diella was elevated to cabinet minister status in September 2025—including delivering a “speech” in parliament and being promoted internationally—no new agreement or additional consent was sought. Bisha regarded this as exploitation of her identity and a violation of Albania’s personal data protection law, which has been aligned with European Union standards.
The personal impact on Bisha has been significant. She reported experiencing harassment in public spaces and online; people called her “Minister Diella,” and some citizens who opposed the government redirected their hostility toward her. Furthermore, Bisha claimed that AKSHI had filed a patent for the avatar without her knowledge, potentially restricting her professional use of her own identity in the future. In her lawsuit, she requested temporary suspension of the use of her face and voice as well as moral damages of one million euros. Her lawyer, Aranit Roshi, emphasized that sanctions for personal data violations by state institutions could reach 21 million euros.
On 23 February 2026, the Tirana Administrative Court rejected the request for interim measures. Consequently, Diella continued to operate while the main legal proceedings remained ongoing. The Albanian government dismissed the lawsuit as “nonsense” but stated its readiness to resolve the matter in court.
The Bisha case sharply exposes failures of ethical governance in AI development. Contractual transparency, respect for individual rights, and clear consent mechanisms were disregarded in favor of political image-building and rapid implementation.
4.3 The Corruption Scandal at AKSHI: The Anti-Corruption Paradox That Undermines Credibility
The deepest irony emerged from within the very institution that created Diella. In 2026, Albania’s Special Anti-Corruption Unit (SPAK) opened an investigation into AKSHI leadership over allegations of massive corruption. According to an OCCRP investigative report published in August 2026, senior AKSHI officials were alleged to have manipulated tender processes to benefit groups linked to criminal organizations. They were accused of receiving kickbacks in various forms, including cash hidden in wine boxes, and of involvement in intimidation practices and, in some reports, kidnapping related to contract disputes.
Authorities ordered the seizure of assets valued at more than 40 million euros, encompassing real estate, vehicles, bank accounts, and cash. The director and deputy director of AKSHI were linked to violations of public procurement rules and cooperation with criminal networks. As of August 2026, the investigation was still ongoing and not all parties had been formally charged; nevertheless, the scale of the asset seizures indicated the complexity and depth of the case.
The resulting paradox is sharp and difficult to refute: an AI system explicitly designed and promoted as an instrument to “eliminate 100 percent of corruption” in public procurement was itself born and managed by an institution under investigation for systemic corruption in procurement. This empirically demonstrates that the quality and integrity of an AI system cannot be separated from the integrity of the humans and institutions that design, train, and operate it.
4.4 Analysis: Algorithmic Transparency as a Concrete Manifestation of Human Intelligence
The Diella case as a whole affirms the central thesis that artificial intelligence is highly dependent on human intelligence. This dependence is not abstract; it is concretely realized through three main pillars:
First, algorithmic transparency.
To date, there has been no adequate public explanation of how Diella evaluates tender documents, establishes criteria, or generates winner recommendations. The absence of explainability and auditability creates a dark space vulnerable to manipulation. Without transparency, the algorithm ceases to be an objective tool and becomes a potential extension of hidden interests.
Second, accountability.
When decisions are transferred to an AI system, a fundamental question arises: who bears legal and political responsibility if errors, bias, or harm occur? In the Diella case, this lack of clarity is further complicated by its still-ambiguous legal status (the Albanian Constitution requires that ministers be natural persons).
Third, ethical and institutional oversight.
The failure to respect individual rights (the Anila Bisha case) and the failure of internal controls culminating in alleged corruption at AKSHI demonstrate that without critical, integrity-driven, and continuous human oversight, AI cannot correct systemic weaknesses. On the contrary, AI reproduces and even legitimizes those weaknesses under the guise of advanced technology.
Thus, Diella is not merely a technical or local political failure in Albania. It constitutes sharp empirical evidence that without human intelligence expressed through algorithmic transparency, clear accountability, and strict ethical oversight, artificial intelligence will never become a solution; it will remain only a mirror reflecting—and at times magnifying—the shortcomings of the humans who create it.
5. Conclusion
The Diella case in Albania—from its launch as a virtual assistant in January 2025, its appointment as the world’s first AI minister in September 2025, to the emergence of Anila Bisha’s personal data violation lawsuit and the large-scale corruption scandal at AKSHI—offers highly valuable empirical lessons. Diella has become not only a symbol of a nation’s digital ambitions but also a clear mirror revealing the limits of artificial intelligence.
Artificial intelligence is one of the most valuable technological achievements in modern human history. It can process data at a scale and speed far beyond individual human capacity, improve the efficiency of public services, accelerate data-driven decision-making, expand access to information, and open opportunities for innovation across many domains of life—from education and health to governance. In the appropriate context, AI can serve as a highly beneficial instrument for enhancing quality of life, reducing administrative errors, and even helping to diminish opportunities for routine and procedural corruption.
However, these substantial benefits can be realized only if human intelligence remains placed above AI. The Diella case clearly demonstrates that AI possesses neither moral consciousness, intrinsic integrity, nor the capacity to correct itself from biases or abuses originating with the humans who create it. When algorithmic transparency is neglected, when accountability is obscured, and when ethical oversight is weakened, AI risks becoming a tool that reproduces—and even legitimizes—human weaknesses and narrow interests.
Human intelligence, with all its limitations, retains irreplaceable advantages: the capacity to reflect on values, to distinguish right from wrong morally, to consider long-term consequences, and to take responsibility for decisions before fellow humans and history. AI is merely an extension of that intelligence, not its replacement. It is a tool, not a ruler.
Therefore, in every effort to integrate AI—whether in government, education, or daily life—the fundamental principle must be upheld: humans remain the primary subjects, while AI is the object that is served and supervised. Algorithmic transparency, critical human oversight, and robust ethical frameworks are not optional supplements; they are absolute prerequisites for AI to become a genuine blessing rather than a concealed threat.
Diella, originally expected to be the “sun” illuminating Albania’s anti-corruption path, ultimately highlighted the dark shadow of technology’s dependence on the moral and intellectual quality of human beings.
This lesson is universal. Artificial intelligence is highly beneficial to human life, but only so long as human intelligence—with its full reflective, ethical, and responsible capacities—remains above it.
Last but not least, Allah SWT has reminded us in His words in QS. Al-‘Ankabūt [29]: 43:
وَتِلْكَ الْأَمْثَالُ نَضْرِبُهَا لِلنَّاسِ ۖ وَمَا يَعْقِلُهَا إِلَّا الْعَالِمُونَ
“And these examples We present to the people, but none will understand them except those of knowledge.”
This verse reminds us that although AI can process millions of data points and produce “conclusions,” true understanding (ya‘qiluhā) remains the exclusive possession of knowledgeable and reasoning human beings.
References
1. MYR Agung Sidayu. (2024). A Historical and Philosophical Analysis of the Modern Philippine Educational System: Legal Foundations and Policy Reforms.
2. Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., … & Vayena, E. (2018). AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations. Minds and Machines, 28(4), 689–707.
3. Mittelstadt, B. D., Allo, P., Taddeo, M., Wachter, S., & Floridi, L. (2016). The ethics of algorithms: Mapping the debate. Big Data & Society, 3(2), 1–21.
4. Organized Crime and Corruption Reporting Project (OCCRP). (2026, 11 August). Albania’s Cheerful ‘AI Minister’ Is the Product of a Government Agency Under Investigation for Massive Corruption. https://www.occrp.org
5. Reuters. (2026, 13 February). Albanian actor sues government for using her image as “AI minister”. https://www.reuters.com
6. Taylor, A. (2025, 11 September). Albania appoints world’s first AI-made minister. Politico Europe. https://www.politico.eu
7. Delauney, G. (2025, 12 September). World’s first AI minister will eliminate corruption, says Albania’s PM. BBC News. https://www.bbc.co.uk
8. Wikipedia. (2026). Diella (AI system). https://en.wikipedia.org/wiki/Diella_(AI_system
9. Associated Press / AP News. (2025–2026). Various reports related to the appointment and controversies surrounding Diella in Albania.
10. Balkan Insight. (2025–2026). Investigative reports and political reactions to Diella as AI minister.
11. The New York Times. (2026, 27 January). Albania Created an ‘A.I. Minister’ to Curb Corruption. Then Its Developers Were Accused of Graft.
12. DW (Deutsche Welle). (2025–2026). Coverage of Diella and the Anila Bisha lawsuit.
13. CNA.al & Voxnews.al. (2026, February). Reports on the Tirana Administrative Court ruling regarding Anila Bisha’s lawsuit.
14. Government of the Republic of Albania / Council of Ministers. (2025). Official documents concerning the establishment and mandate of the Minister of State for Artificial Intelligence “Diella”.
15. National Agency for Information Society (AKSHI). (2025). Public information on the launch and development of Diella on the e-Albania platform.