Albania has appointed 'Diella', the world's first AI minister, to oversee public procurement and fight corruption. The move, while aimed at transparency, has sparked a constitutional debate and raised serious questions about accountability and the future of AI in governance.
In a move that feels pulled from the pages of science fiction, the Albanian government has appointed an artificial intelligence to its cabinet. Prime Minister Edi Rama introduced Diella, the world's first AI-generated minister, in a bid to tackle one of the country's most persistent problems: corruption in public procurement. This decision places Albania at the forefront of a global experiment in governance, raising profound questions about the role of technology in politics and the very definition of a public servant.
Diella, whose name means 'sun' in the native Shqip language, is more than just lines of code. She is presented as a virtual woman dressed in a traditional Albanian folk costume from the Zadrima region. Developed in partnership with Microsoft, Diella is intended to function as what Prime Minister Rama described in a Facebook post as “a member of the cabinet who is not present physically but has been created virtually”.
Her promotion to a ministerial role is a significant step up from her previous duties. Before her cabinet appointment, Diella was already a functional AI-powered virtual assistant on the e-Albania platform, where she assisted citizens in obtaining official documents and navigating government services.
Diella's primary mandate is to bring transparency and impartiality to the process of awarding government contracts. Historically, public procurement in Albania has been susceptible to bribery, favoritism, and vested interests. The government's hope is that by automating key aspects of the decision-making process, it can remove the human element that allows corruption to flourish.
The AI bot will follow a step-by-step process to weigh every tender on merit, playing a significant part in selecting the winners of public contracts. The goal is to prevent government ministries from unfairly influencing outcomes. Prime Minister Rama is confident in the system's potential, stating, “Not only will we wipe out every potential influence on public biddings – we will also make the process much faster, much more efficient and totally accountable.” For a country actively seeking accession to the European Union, this is a strong signal that it is serious about meaningful reform.
The appointment has not been without its detractors. The country’s opposition Democratic Party swiftly condemned the decision, labeling it “ridiculous” and “unconstitutional.” The core of their legal argument rests on Albania's constitution, which stipulates that government ministers must be “mentally competent citizens” aged 18 or over—a definition a virtual entity cannot meet.
Beyond the constitutional debate, critics have raised several practical and ethical concerns:
While Diella is the first AI to be given a ministerial title, Albania is not alone in integrating named AI systems into its government. The United Kingdom launched a multi-tool suite called Humphrey, whimsically named after a fictional bureaucrat from the sitcom Yes Minister. In France, an AI system named Albert was introduced to help public officials answer frequently asked questions, transcribe legal hearings, and streamline administrative tasks. However, Diella's role in direct, high-stakes decision-making represents a significant escalation of AI's role in governance.
While citizens cannot directly consult Diella on public tenders, her origins lie in public service. The underlying technology and virtual assistant are part of the e-Albania platform, the country's central portal for digital government services. Albanian citizens can interact with her virtual assistant persona there to get help with official documents and other public services.
Albania's appointment of Diella is a bold and polarizing step. Supporters see a new era of clean, efficient, and data-driven governance. Critics see a constitutional overreach and a dangerous abdication of human responsibility. The success of this experiment will depend heavily on establishing legal clarity, ensuring robust human oversight, and building public trust. The world will be watching to see if Diella becomes a model for transparent government or a cautionary tale about the perils of moving too far, too fast.
So people who couldn’t solve corruption are now hiring a machine to do it for them. Isn’t this just shifting responsibility to an impersonal algorithm that can’t defend itself or explain its decisions in court? It seems like the perfect scapegoat. If something goes wrong, “the AI made the mistake.” If it works, “we implemented the AI.” Convenient. What if a rule is secretly inserted into the system to “prefer companies with a blue logo”? Who will check it if the code is a trade secret?
But what if this is just the beginning? Imagine an AI mayor who allocates budgets for road repairs not based on who shouts the loudest, but based on data on traffic density and road conditions. Or an AI judge for minor offenses who is completely impartial and uninfluenced by whether you have an expensive car. How about AI in real estate that would detect cartels and artificial rent increases in real time? Diella may only assess contracts, but her descendants could manage entire city ecosystems, from transportation to energy, with a single metric: maximum efficiency and fairness for citizens. Maybe you’re not afraid of AI failing, but of it succeeding and showing how inefficient you’ve been.
