B(IA)S
Algorithms in Power: Democracy in the Age of AI Candidates
by Alessandro Mancini
New technologies are changing our world forever. The question is: for the better or for the worse?
What are the risks, the shadows, the dangers?
In 2027, Rome may witness a historic—and perhaps slightly surreal—turning point: its first mayoral candidate entirely generated by artificial intelligence. Her name is Francesca Giubelli, a virtual influencer complete with pictures of pets, airplane trips, AS Roma fan gear, and the inevitable #ADV posts.
Giubelli has recently become the face of the civic list AI
Roma, which promises a revolution in city governance through a more efficient, transparent, and participatory model. Behind this digital figure is a team of Italian developers who have entrusted AI systems with defining the candidate’s identity, electoral platform, and communication strategy—a pioneering experiment on the European political scene.
The project was born from the ambition to overcome the inefficiencies and shortcomings of traditional politics, offering a “clean” figure, free from scandals, capable of analyzing data in real-time and engaging with citizens 24/7, without lapses in attention or emotional manipulation.
For its creators, this is more than a tech gimmick: it represents a proposal to open what they describe as a new democratic frontier, where the line between real and virtual becomes increasingly blurred. By using these tools, they also aim to provoke, much like what is happening in Taranto with Anna Luce D’Amico, the first Italian mayoral candidate created entirely with generative AI. Conceived by a team of creatives and technologists, this virtual candidate is a provocative, experimental project that challenges the actual ability of current politics to listen to, understand, and solve citizens’ problems. Similarly, in Rende, Calabria, for the 2025 municipal elections, an unusual candidate has emerged: Gennaro Chiappetta, nicknamed “U sindacu,” a digital avatar who takes part in the campaign by interacting in real time with voters and journalists on social media.
However, not all such experiments are driven by noble intentions. A recent investigation by Check Point Research revealed that at least ten countries have used AI in recent elections to produce persuasive, manipulative, or even deceptive content—disseminated especially in the days just before voting. These strategies include auto-generated speeches, managed social campaigns, and even deepfakes used for self-promotion or attacking opponents.
A striking example is that of Deputy Prime Minister Matteo Salvini who, during last year’s European elections, used generative AI to create and share campaign posters and political slogans on social media with derogatory and provocative content targeting the LGBTQ+ community and migrants—all without disclosing that the content was AI-generated.
Italy is not alone in exploring this frontier. Around the world, political AI candidates and AI-powered campaign communication are on the rise. One of the best-known examples is AI Steve in the United Kingdom, the AI alter ego of 2024 parliamentary candidate Steve Endacott, powered by a generative AI system designed to interact directly with voters and develop a real-time platform based on their input. In India, during the 2024 campaign, Senthil Nayagam and his startup Muonium used generative AI to produce realistic videos featuring the appearance and voice of famous deceased politicians, such as Rajiv Gandhi, Indira Gandhi, Narasimha Rao, and Atal Bihari Vajpayee. “We’re using AI, but with good intentions,” Nayagam said. “We don’t want to abuse the technology or create deepfakes.”
In neighboring Pakistan, former Prime Minister Imran Khan used AI last year to campaign from prison, where he was held by political opponents to prevent his party’s re-election. His supporters used generative AI to spread messages on social media and messaging platforms using his likeness and voice. This allowed voters to counter official propaganda and the government-imposed ban on rallies, ultimately helping Khan win the election despite his physical absence.
All these cases point to a transformation that affects not only how political communication is conducted but also the very substance of the relationship between representatives and the represented.
While digital politics increasingly turns to AI and deepfakes, pop culture had already anticipated these dynamics. In the well-known 2023 Black Mirror episode The Waldo Moment, a digitally animated character born as satire—Waldo, a cheeky blue bear—runs in a political election and wins over voters with his blunt, provocative style, far removed from the clichés of traditional politics. That story, almost prophetic today, foreshadowed both the risks and opportunities of technology-mediated political communication, warning about manipulation mechanisms and the erosion of trust between citizens and institutions.
The arrival of AI-generated political candidates raises profound questions about what it means to participate in democracy today. On the one hand, AI can make campaigns more effective and personalized, providing clear and up-to-date information. On the other hand, it risks depersonalizing political discourse, turning debate into a field dominated by algorithms and digital marketing strategies. Davide Bennato, Professor of Digital Media Sociology at the University of Catania, believes these candidacies are a mirror of our expectations—and disillusionment—with human politics: “Has politics become just an empty exercise in rhetoric at the expense of civic engagement?” the professor asks. “If an AI can convincingly simulate a political platform […] and the associated tone of campaign messaging, this could indicate that political representation has become a collection of clichés incapable of truly engaging citizens.”
In this scenario, transparency becomes essential: voters must know when they are interacting with AI, what data is used to shape messages, and which interests are behind the campaigns. Updated and strict regulation is also needed to combat the spread of fake news and manipulative content that could undermine institutional credibility and the freedom of choice.
On the other hand, as the magazine Magia News points out, AI can offer a chance to reconnect citizens with a political system perceived as distant, by creating new spaces for participation and debate. Imagine candidates or political systems capable of listening to millions of citizens in real time, formulating proposals based on objective data, and encouraging public discourse without ideological or lobbyist filters. If well governed, this vision could lead to a more inclusive, efficient politics that truly serves the community—with all the inherent limits and risks.
One thing is certain: the phenomenon of artificial intelligence in politics is no longer a distant prospect—it is an emerging reality that urges us to reflect on the delicate balance between technological innovation, ethics, and democratic participation. The line between real and virtual is increasingly blurred, and the challenge lies not only in adopting new technologies, but above all in governing them to preserve the quality of political discourse and citizens’ trust in institutions.
Alessandro Mancini
Is a graduate in Publishing and Writing from La Sapienza University in Rome, he is a freelance journalist, content creator and social media manager. Between 2018 and 2020, he was editorial director of the online magazine he founded in 2016, Artwave.it, specialising in contemporary art and culture. He writes and speaks mainly about contemporary art, labour, inequality and social rights.
