On September 11, the head of the Albanian government, Edi Rama, appointed a minister generated by an artificial intelligence system. Named Diella (which means sun in Albanian) and equipped with an avatar dressed in traditional costume, she is responsible for awarding public contracts. The stated aim is to combat corruption. According to the Albanian Prime Minister, ” every public penny will be allocated in a perfectly transparent manner “.
Source: Génétique et Matière à réflexion
Albania’s new AI-generated minister makes her first speech to parliament
.
The first minister to be born of artificial intelligence said that she was “not there to replace people, but to help them”.
But ” this humanisation is a political strategy, not a technological advance “, says Roxana Ologeanu-Taddéi, lecturer in digital management at the University of Montpellier and author of the book Intelligence artificielle et anthropomorphisme. De l’illusion à la confusion (from illusion to confusion)
” By disguising a machine as a conscience, power erases its own responsibilities and transforms the illusion of objectivity and progress into an instrument of legitimisation “, she points out. A study just published in Nature[1] shows that our moral brakes are further weakened when people delegate tasks to AI. It is already ” easier ” to bend or break the rules when no one is looking, or when someone else is doing the deed.
“Anthropomorphism has consequences”.
But ” this confusion is not insignificant“, warns the researcher. When we accept the idea that a chatbot can love or suffer, we blur the boundary between humans and machines in the collective imagination“. And ” anthropomorphism has consequences “: ” the real risks are political, economic and environmental “.
Indeed, ” generative AI consumes huge amounts of energy, exacerbating climate pressures. It concentrates power in the hands of a few companies that control data, infrastructures and platforms. It is reconfiguring labour markets, de-skilling some professions and overloading others. It amplifies bias and misinformation, with profound implications for democracy ” (cf. Development of artificial intelligence: “health is being used as a pretext”).
The risk of infantilisation
Beyond that, ” anthropomorphism infantilises us “, says Roxana Ologeanu-Taddéi. Because ” it leads us to believe that machines could take care of our problems, share our concerns or relieve us of our responsibilities “. But machines don’t care,” she points out. They have no interest in the climate, in inequality or democracy. These issues are and will remain ours.
Sources for the press summary: La Croix, Roxana Ologeanu-Taddei (16/09/2025); Ouest France, Cécile Réto (15/09/2025); Phys.org, Max Planck society (17/09/2025)
[1] Nils Köbis, Delegation to artificial intelligence can increase dishonest behaviour, Nature (2025). DOI: 10.1038/s41586-025-09505-x
