The 61st Venice Art Biennale closes on 22 November with a split verdict on AI. One art historian called it AI's first major victory at a large international exhibition. Other critics saw a Biennale that kept a deliberate distance from the technology.
Both readings hold, because they describe different parts of Venice. The curated main exhibition barely touches the subject. The machines are in the national pavilions, the collateral events and the shows around town. This roundup covers where to find them and what they say about art and people in the age of AI.
A main exhibition tuned away from the machine
The central show, In Minor Keys, runs from 9 May to 22 November 2026 at the Giardini and the Arsenale. It was conceived by Koyo Kouoh, who died in May 2025, and realised by the team she had chosen. It gathers 111 artists and collectives, many from the Global South.
Its brief is to slow down and listen: to quiet voices, to craft, to repair. Generative AI is not part of that vocabulary. Reviewing the opening, Daily Sabah read the show's distance from AI as its clearest signal: the faster the technology moves, the more culture reaches for the body, ritual and the handmade. The most talked-about works were physical, from Florentina Holzinger's performances in the Austrian Pavilion to the dolls visitors carry through the Japanese one.
That silence is itself a position on AI. The answer to the question arrives from the 100 national participations and 31 collateral events around the main show, where each country and institution sets its own agenda.
The pavilions: where the machines are
Only a handful of the 100 national participations engage with machines directly, and most use technology as a tool rather than a subject.
| Pavilion | Artist / work | Venue | What it does |
|---|---|---|---|
| Venice | Paolo Fantin and Dardust with Cisco and H-Farm, Echo, within Note persistenti | Padiglione Venezia, Giardini | Visitors share an impression of the city and an AI system turns it into light, music and motion. The most openly corporate work in the Biennale; much of its coverage comes from the partners. |
| Taiwan (collateral event) | Li Yi-Fan, Screen Melancholy | Palazzo delle Prigioni | A 60-minute video built with game engines, animation and generated images: puppet characters, deformed bodies, absurd digital rooms. Domus called it a video game gone terribly wrong. |
| Portugal | Alexandre Estrela, RedSkyFalls | Fondaco Marcello, San Marco | A cybernetic installation that reacts in real time, with sound and image, to earthquakes anywhere on the planet. Replicas in Lisbon and Los Angeles respond to the same tremors. |
| Denmark | Maja Malou Lyse with DIS, Things to Come | Giardini | Film and installation on falling male fertility, built around a study claiming that virtual-reality pornography raises sperm motility. Technology as subject, not tool; Kunstkritikk was sceptical. |
| Zimbabwe | Five artists, Second Nature | Manyonga, curated by Fadzai Veronica Muchemwa | Castello | Mostly physical works on human habits now entangled with machine logic. Art about AI, made by hand. |
| Cyprus | Xenofontos, it rests to the bones | Venue not confirmed | Uses anti-spectacular machines, motors and automata to make data, debris and memory accumulate and dissolve. The humblest robots in Venice. |
| Finland | Jenna Sutela, Aeolian Suite | Aalto Pavilion, Giardini | Wind data from Venice, Helsinki and elsewhere becomes a live sound piece. Not an AI work, but a non-human system as co-composer. |
| Holy See | Soundwalk Collective and others, The Ear Is the Eye of the Soul | Carmelite garden, Cannaregio, and Santa Maria Ausiliatrice, Castello | An instrument in the garden turns plant bioelectric activity into sound in real time. |
| China | Dream Stream, curated by Yu Xuhong | Arsenale | A robot writes "Dream Stream" in the calligraphic hand of Wang Dongling, next to his monumental Free and Easy Wandering. The pavilion also shows the video game Black Myth: Wukong as art. Artribune counted it among the disappointments of the opening week. |
The Chinese robot is the most literal image of the year: a machine performing one of the most human of East Asian art forms. Portugal and Finland point the other way, treating the planet itself as the system the artist listens to.
Around town: where the AI debate really happened
The sharpest thinking about AI happened outside the Biennale's gates. Artribune put the exhibitions that took on technology among the best things in Venice this year, and The Art Newspaper found collateral shows that gave digital art something to feel good about.
| Exhibition | Venue | Dates | Why go |
|---|---|---|---|
| Strange Rules | Palazzo Diedo, Cannaregio | Until 22 November | The rules behind AI, platforms and biology, treated as the artwork |
| Canicula | Complesso dell'Ospedaletto, Castello | Until 22 November, free entry | Eight new video installations on technology, overload and propaganda |
| Aghrab Idrāk | Palazzo Cavanis, Zattere | Until 22 November | VCUarts Qatar's research labs, including an AI-powered one-eyed robot |
| Painting in the Present Tense | Palazzo Cini, Dorsoduro | Closed 27 September | David Salle paints over the output of an AI model trained on his own work |
| Rage Bait | Palazzo Franchetti and Giudecca | Closed 30 June | Eva & Franco Mattes on outrage, platforms and AI data scraping |
Strange Rules
Curated by artists Holly Herndon and Mat Dryhurst with Hans Ulrich Obrist and Adriana Rispoli, this is the essential stop. It introduces "Protocol Art": work that treats the rules behind algorithms, AI and platforms as the material. Herndon told Ocula that they did not want it to read like an AI show, and the highlights bear that out:
- Holly Herndon & Mat Dryhurst, Attention Guild (2026): described as a parliament of humans and agents, built with the architecture studio SUB.
- Trevor Paglen, Voyager (2026): a hypnosis session led by an AI-generated voice that responds to the visitor's answers and heart rate. There is no screen at all.
- Michael Levin, Double-Headed (2026): live flatworms that the Tufts biologist persuaded, through bioelectricity, to grow two heads. The curators call them the show's mascots: a fork in the road for where technology is taking us.
- Ken Stanley, Picbreeder (2008–2020): images bred collectively with an early generative system by the AI researcher.
- Fabien Giraud, The Feral – Epoch 1 (2025–2026): a video made with machine learning and generative AI, the first chapter of a project meant to run until 3025.
- Also on show: Lynn Hershman Leeson, Philippe Parreno, Stephanie Dinkins, Ayoung Kim, Agnieszka Kurant, Ho Tzu Nyen and Lawrence Abu Hamdan.
Canicula
Fondazione In Between Art Film commissioned eight video installations from ten artists for Canicula, curated by Alessandro Rabottini and Leonardo Bigazzi. It closes the foundation's "Trilogy of Uncertainties", after Penumbra (2022) and Nebula (2024). The title refers to the dog days of summer, used as a metaphor for a culture overloaded by technology, inequality and propaganda.
It is not an AI show, but five works speak directly to what machines do to images, attention and power:
- Yuyan Wang, Boring Billion (2026): found online footage of industrial machinery and humanoid robots. Machines are built to mirror living systems, while humans adapt to mechanised logic.
- P. Staff, Terminal Lucidity (2026): an endless film whose segments an algorithm reshuffles at random, so no two visits are the same.
- Massimo D'Anolfi and Martina Parenti, 24 Landscapes + A Vision (2026): three screens on technology's role in destruction and care, plus a thermal camera that films the visitors live. The audience becomes the watched.
- Wang Tuo, The Experimental Paradigm of Ownership and Autonomy (2026): a fiction built partly from mass-media, surveillance-camera and propaganda footage.
- Lawrence Abu Hamdan, 450XL: The Story of a Fugitive Sound (2026): his research agency Earshot reconstructs a sonic weapon allegedly used against a silent vigil in Belgrade in March 2025.
The complex is open Wednesday to Monday, 10:00 to 18:00, and free guided tours run on the second and fourth Saturday of each month.
Aghrab Idrāk, Palazzo Cavanis
An official collateral event by VCUarts Qatar, built by research labs rather than single artists. At its centre is Mesh Lab's A Monocular Monologue: a 3D-printed, single-eyed robot that fixes its gaze on visitors and whispers a monologue placing AI in the lineage of ancient myth. The Art Newspaper compared it to a modern-day doomsayer. Boost Lab's Chrysalis is a breathing sculpture that filters the air, and xLab's Preceding Emptiness treats Arabic script as a living system in kinetic light.
Closed, but worth knowing
David Salle, Painting in the Present Tense. Salle trained a custom AI model on his own Tapestry Paintings of 1989–1991, printed its output on canvas and painted over it by hand. In an interview with Finestre sull'Arte he describes training the system on his own visual archive, and his press release puts the strategy in one line: "One way to subvert a hegemonic technology is to co-opt it." The works have since travelled to Paris.
Eva & Franco Mattes, Rage Bait. Installation, video and generative AI showed how outrage becomes the natural output of platforms optimised for engagement. The Art Newspaper noted its focus on AI data scraping; Artribune called it probably the fiercest project in Venice on the emotional capitalism of social platforms.
What it adds up to
Venice 2026 shows five distinct answers to AI, and they rarely share a room.
| Stance | What the artist does | Where to see it |
|---|---|---|
| AI as instrument | Uses a model or robot as one more tool in the studio | Salle, Echo in the Venice Pavilion, China's calligraphy robot, Giraud in Strange Rules |
| AI as subject | Makes work about life under algorithms, often by hand | Taiwan, Denmark, Zimbabwe, Canicula, Rage Bait |
| AI as system | Treats the rules and protocols as the artwork | Strange Rules |
| Other intelligences | Treats wind, worms or earthquakes as co-authors | Portugal, Finland, the Holy See, Levin's worms in Strange Rules |
| Refusal | Answers with the body, craft and slowness | In Minor Keys, the Austrian and Japanese pavilions |
The "first major victory" claim needs a caveat. Liivrand counted Hong Kong and Macao among the AI pavilions. Their own descriptions point elsewhere: light, sound and everyday objects for Hong Kong, a tribute to the Qing-dynasty painter Wu Li for Macao. AI is present in Venice, but concentrated in a handful of venues.
The recurring question is where the art sits when a machine can reproduce the gesture. A robot can write in Wang Dongling's hand, yet the style it copies took him a lifetime. Salle's model generates endless variations, and he still chooses and paints. In both cases the value moves from execution to judgement: what to ask for, what to keep, what to refuse.
The second question is about people, not artists. The works that stay with visitors are the ones that show what the technology does to attention, memory and autonomy: Paglen's voice steering a mind, the thermal camera that turns Canicula's visitors into its subject, the Mattes' outrage machine, Zimbabwe's habits rewired by machine logic. The main exhibition's reply is to hand visitors a doll to carry. Both are responses to the same pressure.
For those of us who build AI and robotic systems for a living, the lesson is practical. The convincing works keep the human decision visible and treat the model as infrastructure. The weak ones put the technology on a plinth and call it the future.
If you go before 22 November
The Biennale, its pavilions, Strange Rules, Canicula and Aghrab Idrāk all close on 22 November 2026; Canicula is free. The Salle and Mattes exhibitions have already closed. Check opening days on the official Biennale site before travelling.
Sources and further reading
- La Biennale di Venezia: Biennale Arte 2026, In Minor Keys, Portugal Pavilion and Holy See Pavilion
- Biennial Foundation: national pavilions and collateral events, full list
- ERR News: Harry Liivrand on AI's first major breakthrough
- Daily Sabah: Venice Biennale 2026, pavilion of reality
- Artribune: the best and worst of the 2026 Art Biennale (in Italian)
- The Art Newspaper: collateral exhibitions and digital art and the curatorial team on In Minor Keys
- China Academy of Art: Dream Stream at the China Pavilion; Rocking Robots: the calligraphy robot
- Domus: pavilions and exhibitions not to miss and exhibitions across the city
- Cisco Newsroom: Echo at the Venice Pavilion; Exibart: the Venice Pavilion project (in Italian)
- Artforum: Alexandre Estrela on RedSkyFalls and Maja Malou Lyse on Things to Come; Kunstkritikk: review of the Danish Pavilion
- e-flux: Strange Rules and Rage Bait; Ocula: Holly Herndon and Mat Dryhurst interview
- Fondazione In Between Art Film: Canicula digital guide
- designboom: VCUarts Qatar, Aghrab Idrāk and guide to the national pavilions
- Finestre sull'Arte: interview with David Salle; Thaddaeus Ropac: press release
- NewsDay Zimbabwe: Second Nature | Manyonga