In August 2022 the annual fine-arts competition of the Colorado State Fair awarded first prize in 'Digital Art / Digitally Manipulated Photography' to Jason M. Allen. The work, *Théâtre D'opéra Spatial*, depicts figures in glowing costumes on a vast stage gazing into a starlit window — a neo-baroque scene. Allen had not painted it; he had written hundreds of text prompts to Midjourney — then only a few months old — selected outputs, retouched in Photoshop, and printed the result onto a large canvas.

The jury did not know AI was involved — the entry read 'Jason M. Allen via Midjourney', but Midjourney was not widely recognised in the art world as a software tool at the time. First prize carried $300. When the result reached social media, the debate exploded within hours: artists pointed out that their own work had been used without permission to train such models, the jury refused to revoke the prize, and Allen replied that 'this was a competition I entered and won; this was my work'.

The event marked a rupture far larger than a single state fair. In mid-2022 Stable Diffusion, DALL-E 2 and Midjourney went public almost simultaneously; generative AI had crossed a usability threshold in image production. Allen's win was the first clear demonstration that an AI-generated image could be entered into professional competition with human artists, judged by humans, and prevail.

The US Copyright Office subsequently refused to register the work — and many AI-generated works after it — on the grounds that they lacked sufficient human authorship; legally, *Théâtre D'opéra Spatial* cannot be copyrighted. This called the very concept of authorship into question: is the person who writes the prompt, selects among hundreds of outputs, and applies the final retouching the 'author'? Every debate about AI art after 2022 unfolded in the shadow of this single entry.