The painting hangs in a gilt frame on the third floor of the Whitney Museum, a luminous landscape of impossible colours that seems to shift as the viewer moves. It was created in eleven seconds by a generative AI model trained on four hundred years of Western art. And it has started a war.

Since the Whitney announced its decision to include three AI-generated works in its spring exhibition, the response from the art world has been swift and sharply divided. A coalition of more than two hundred artists signed an open letter calling the decision a betrayal of human creativity. Meanwhile, a smaller but vocal group of technologists and digital artists has praised the museum for recognising a new medium.

The debate is not confined to New York. The Tate Modern in London recently rejected a donated AI work, while the Centre Pompidou in Paris has commissioned an entire wing dedicated to computational art. In Tokyo, the Mori Art Museum has adopted a middle path, exhibiting AI works but labelling them explicitly as machine-generated.

"The question is not whether AI can produce beautiful images," said Clara Montoya, a curator at the Museum of Contemporary Art in Barcelona. "The question is whether beauty alone constitutes art. Art requires intention, context, struggle. A machine has none of these."

Proponents counter that the human role in AI art is more substantial than critics acknowledge. Selecting training data, crafting prompts, and curating outputs all involve creative judgment, they argue. The AI is a tool, no different in principle from a camera or a printing press.