nobody. no signature. No passport. Yet, in the small Balkan country of Albania, the “Dila,” or “Sun,” is officially a cabinet minister. Originally a digital assistant on the government’s e-Albania portal, Deila was promoted to the world’s first virtual artificial intelligence minister by Prime Minister Edi Rama in September.
Dressed in folkloric Albanian garb and powered by algorithms, Deila’s AI avatar smiled from a government monitor. Ram promised a new era in which “public tenders will be 100% flawless, and every public fund will be 100% transparent.” In a country that has long struggled with corruption, this promise sounded familiar. But behind Deila’s digital smile stands a very human question: What happens to responsibility when power has no human face?
When President Bajram Begaj approved the new 16-member Cabinet on 15 September, Deila was missing. The virtual minister announced with much fanfare by Ram had no line in the official document. Instead, Article 2 of the decree assigned Rama himself “responsibility for the establishment and functioning of the Virtual Minister”, effectively placing the system under his direct authority.
In practice, this Act gave Ram control over an entity that has no legal existence. Under Article 100 of the Constitution of Albania, each member of the Council of Ministers must be a natural person, capable of deliberating, voting, and bearing both moral and political responsibility.
“The concept of an ‘AI minister’ has no basis in the Constitution of Albania,” said legal expert Sokol Hadjijaj. “The constitutional meaning of the word minister is inseparable from that of a physical person and the responsibilities attached to that role. A minister is accountable to the citizens – something that no algorithm can do.”
Hajizaj said laws on personal data protection and civil service only touch peripheral issues of responsibility, without defining how AI might work or who is legally accountable for its actions.
‘Just a chatbot’
Since Deila’s debut as minister, the government of Albania has not provided any details about the AI’s training data, underlying code, or human supervision. On the e-Albania portal, Diala still appears as a smiling avatar greeting users with simple answers, but there is no evidence that AI has a true decision-making role.
“At the moment, Diala is just a chatbot, not an autonomous system,” said Besmir Semanaj, who has 17 years of experience in information technology. “Artificial intelligence can support government decisions if properly trained and monitored, but the real issue is transparency: we don’t know what data it relies on or who is responsible for maintaining it.”
Adopted in 2024 as the world’s first comprehensive regulation on artificial intelligence, the EU’s AI Act classifies such systems as high risk. Every algorithm affecting public administration in the EU must be reviewed and signed off by a human decision-maker – a safeguard against what the EU calls “automation without accountability”.
As a candidate for EU membership, Albania is not yet bound by these standards, but the government has promised to align the legislation with EU laws.
“If we create systems powered by artificial intelligence, we also have to create institutions to monitor and control them,” Semanaj said. “Investments in AI must go hand in hand with investments in oversight, otherwise we risk creating systems we cannot monitor.”
A political distraction?
When Albania’s new parliament sat on 18 September to present the government’s program for a fourth term, tempers immediately flared. What was meant to be a policy debate ended within minutes of shouting, desk banging and a walkout. As Rama, who holds direct responsibilities as Virtual AI Minister, paused to give the floor to Deila, the chamber’s screens lit up.
A digital figure appeared in traditional Albanian attire, speaking in a quiet synthetic voice – a mix of folklore and code. “Judge me not by my origins, but by my actions,” Deila told lawmakers. “I may not be human, but I am constitutional. I serve the people who wrote the Constitution.”
Moments later right-wing opposition lawmakers, who dispute the validity of elections that determine the composition of parliament in May, walked out in protest.
Artan Fuga, professor of communications and member of the Academy of Sciences of Albania, said that Deila was less a technological milestone than an act of political distraction. “Ram used artificial intelligence to create a second center of attention,” he said. “Instead of debating the government’s legitimacy or programme, the opposition found itself debating an avatar.”
Fuga said the risks, however, were not theatrical. “At a time when Albania still questions whether its elections are truly free, we are told that there is no longer a need to control the government through parliament, express citizen will or demand ethical transparency, because pure intelligence can do it all,” he said.
Beyond the political experiment lies a deeper question: who ultimately decides? Technically, Deila may one day make decisions in an official capacity. But Albanians are not convinced that Deila should be given the power to do so.
“Algorithms can process information — but they can’t assess moral consequences,” Fuga said. “Once you elevate technical rationality above political accountability, you risk destroying democracy itself.”