Alessandra Navazio
What is left of a political manifesto if you strip away the message? In AI Manifesta, Francesco D’Isa and Chiara Moresco use Artificial Intelligence to expose the visual grammar of propaganda: hands, fists, flags, symbolic colours. A project that is at once critical investigation, curated archive and visual vaccine against political pathos.
A clenched fist, a raised hand, a flower blooming against a bright red background. Political posters speak to us before saying anything at all. These are images we know so well that they strike us without our noticing. AI Manifesta, a work by Francesco D’Isa and Chiara Moresco, on display for the past three weeks at Via Zaccherini Alvisi 11/2 in Bologna and curated by Sineglossa and Fondazione Gramsci Emilia-Romagna, begins exactly here: with the silent grammar of visual propaganda.
Francesco D’Isa
Trained as a philosopher and digital artist, has exhibited internationally in galleries and contemporary art centres. After debuting with the graphic novel I. (Nottetempo, 2011), he has published essays and novels with Hoepli, effequ, Tunué and Newton Compton. His latest novel is “La stanza di Therese” (Tunué, 2017), and his philosophical essay “L’assurda evidenza” was published by Edizioni Tlon in 2022. His most recent works are the graphic novel “Sunyata” (Eris Edizioni, 2023) and the essay “La rivoluzione” algoritmica delle immagini (Sossella Editore, 2024). He is editorial director of the cultural magazine “L’Indiscreto” and writes and illustrates for various Italian and international publications. He teaches Philosophy at the Lorenzo de’ Medici Institute (Florence) and Illustration and Contemporary Plastic Techniques at LABA (Brescia).
But what happens if we strip these posters of their content? If we leave only the persuasive forms? Is it possible to use Artificial Intelligence not to generate slogans, but to reveal and dismantle the mechanisms that make them effective? In an age of technological scepticism, this collection of 280 political posters generated with AI offers a radical visual investigation, a vaccine against pathos, and a reflection on the role of AI in the arts. Here is how.
The research behind AI Manifesta
The very title of the project is a declaration of intent: at the core of AI Manifesta, by philosopher Francesco D’Isa and artist Chiara Moresco, lies a deliberate philosophical and artistic choice—to use Artificial Intelligence as a lens that manifests, that is, makes visible, the recurring patterns of political propaganda. Over the course of a year, this investigation led to the identification of six persuasive structures within political messaging: hands, faces, writing, flowers, war and flags. Starting from a selection of 100 historical posters from the Fondazione Gramsci, recently digitised1, and using a generative AI, they produced 280 images that appear “real”, but are not: they evoke something familiar, yet say nothing. Since the early days of generative AI, D’Isa has been fascinated by one of its unique abilities: to generate ambiguous, unnameable forms that are nonetheless strangely familiar.
«At first», he recalls, «AI produced images you could not name, but you felt they were “something”. A room full of objects you could not quite identify».
That ambiguity, between form and content, which today is more refined but still present, became in AI Manifesta a tool to reveal what conveys emotion before meaning: the visceral component of rhetoric, understood as the art of persuasive discourse—pathos.
The backbone of rhetorical discourse, as codified by Aristotle in Rhetoric, remains operative today—not only in posters, but across media. Political memes and Instagram carousels, too, rely on the same mechanisms: the construction of credibility (ethos), emotional impact (pathos), and simplified logical argumentation (logos)2.
«Our research revealed that, in posters from the 1950s to today», D’Isa continues, «human figures are often depicted in emotionally intense states: bodies in tension, fists, dramatic gestures, or objects with strong symbolic potential, such as flowers.
«These elements are open to interpretation, evoking multiple meanings yet powerful in their iconic presence. And then there is the composition itself, often structured according to Gestalt principles». The gaze is guided by areas of chromatic and formal contrast, which lead the viewer across the surface of the image to a focal point—usually a phrase or short text.
Posters as “visual vaccines”
The strongest impression left by AI Manifesta is that of ambiguity. Passers-by, say D’Isa and Moresco, often mistake the posters for a genuine political campaign.
«I hope—positively speaking—that the public finds them interesting and falls for them», says D’Isa, deliberately.
«Only to then realise they are meaningless posters, and begin to ask themselves why they were constructed that way», why they were drawn to them. «Even while we were putting them up, an elderly man asked us what we were doing. He was struck by the contrast between the colours and the urban setting. It was something new, something that stood out. And that spontaneous interest is already a result», Moresco recalls.
Chiara Moresco
After her second academic degree at LABA in Brescia in 2025, she continues her education through artistic and editorial collaborations. At the same time, she develops her artistic research focused on the relationship between the body and the concepts of identity, communication and connection with the other. Moresco adopts a precise and multifaceted visual approach, using a variety of languages ranging from installations to video, from sculpture to writing.
The colour of the posters, in fact, cuts across all identified patterns and, while not a category, is essential in making the images stand out. Bright reds, solid colour backgrounds, bold geometries—all contribute to a sense of déjà vu that bypasses rationality, works below the threshold of awareness, and activates what D’Isa and Moresco call the “collective visual unconscious”. Unsurprisingly, the most interesting posters are those that expose the dynamics of pathos most clearly: «the ones with unreadable writing», Moresco explains, «because the eye, in viewing them, follows the compositional flow of a real poster, and once it reaches the expected semantic centre, it finds emptiness».
This is precisely how we can train our eyes to identify persuasive formulas and disarm them. «Exposing the public to a diluted version of a persuasive argument, along with an explanation of the rhetorical trick, triggers a small cognitive conflict that reduces susceptibility to similar messages for several weeks afterwards», explains D’Isa, referring to studies on psychological inoculation3. When the viewer next encounters the same visual cliché, it will be that latent sense of recognition that triggers an inner alert: critical thinking will activate before pathos has a chance to strike.
In this sense, understanding the mechanics of rhetoric is not just about unmasking hostile messages, but also about preserving a critical stance towards those we consider familiar.
The political value of using AI
«Our dream was to spread them all over Bologna», confesses Moresco, because AI Manifesta is not only an exhibition, but also an act of reclaiming—a gesture that restores visibility to an iconographic heritage often confined to archives, and proposes a use of AI with a clear social purpose.
In a context where every generative AI project provokes polarised reactions—some seeing it as a new form of democratisation, others fearing yet another capitalist appropriation—the work of D’Isa and Moresco is not merely an aesthetic experiment, but a deliberate stance expressed through artistic practice. The technique is used in “another” way, one that avoids the logic of performance and blind automation, and instead channels the algorithm’s productive force towards awareness, inquiry and reinterpretation. For it is not the machine that dictates subordination, but the “political and economic direction” that extracts its profits4 and determines its use.
D’Isa states it clearly: he prefers a «proactive attitude», one that shows what can be done with AI, counterbalancing the excess of catastrophic analysis that so often stifles any possibility of thoughtful engagement. It is a way of resisting media panic—that recurring reaction that, with each new technological medium, rushes to blame it for corrupting youth, destroying jobs or upending the truth5.
- Manifestipolitici.it is a project conceived and promoted by the Fondazione Gramsci Emilia-Romagna to enable the consultation and study of paper documents produced to serve an event and not to last over time. ↩︎
- Wang, M., Chang, W., Kuo, K., & Tsai, K. (2023). Analyzing image-based political propaganda in referendum campaigns: from elements to strategies. EPJ Data Science, 12(1). https://doi.org/10.1140/epjds/s13688-023-00407-4 ↩︎
- Roozenbeek, J., Van Der Linden, S., Goldberg, B., Rathje, S., & Lewandowsky, S. (2022). Psychological inoculation improves resilience against misinformation on social media. Science Advances, 8(34). https://doi.org/10.1126/sciadv.abo6254 ↩︎
- D’Isa, F. (s.d.). Chi controlla la macchina. The Bunker. https://www.the-bunker.it/chi-controlla-la-macchina/ ↩︎
- ibidem. ↩︎
