Google and artist Prune Nourry use generative AI to rebuild lost studio spaces of 19th-century women sculptors

Google is partnering with artist Prune Nourry to bring to life the studio spaces of 18 overlooked 19th- and early-20th-century women sculptors in Paris, using generative artificial intelligence as a storytelling medium.

The effort, titled “Visionaries,” opens October 20, 2026 at the Petit Palais in Paris. The exhibition runs through January 30, 2027, with the AI-assisted film installation presented in museum galleries and a companion documentary available through Google Arts & Culture.

Nourry and Google Arts & Culture say the project is designed to counter a long history of erasure in art documentation, where women sculptors were often remembered more as models or muses than as working artists. The exhibition includes monumental plaster portraits sculpted by Nourry, intended to celebrate the women and restore attention to their creative legacies.

Because few historical photographs exist of many of these sculptors’ actual workspaces, the project relies on a mix of archival research and AI reconstruction. The teams say three art historians conducted a year of research using letters, journals, and photographic portraits to inform the digital studio recreations.

Those reconstructions are built into a “walk-through” experience meant to let viewers virtually move through the sculptors’ environments. Google says it used generative AI and image editing models to create 3D versions of the studios and animate camera paths, allowing a controlled perspective of what each workspace may have looked like.

Google Arts & Culture also describes a separate sequence conceived by Nourry and film director Vincent Lorca that uses AI image editing to visually underscore who disappeared from the record. In that scene, female students slowly fade from an archival photograph of Antoine Bourdelle’s studio, leaving only the men, as a way to illustrate how women were sidelined in historical memory.

At the technology level, the company said a custom tool built with Google Gemini was used to create the 3D models from archival photos and to generate the camera movements that animate the virtual studio tours.

Google said the work expands beyond the in-person exhibition through a 10-minute documentary that ties the digital reconstructions to the historical sourcing process. The film, directed by Lorca, is positioned as an explanation of how the team moved from scattered archival materials to a coherent virtual representation of each studio.

Still, the project is explicitly interpretive. The company’s description centers on the difficulty of finding surviving images of actual workspaces and on rebuilding spaces from the available evidence. As presented, the reconstructed studios should be understood as informed creative representations rather than literal photographic records of what the spaces definitively looked like.

Looking ahead, the “Visionaries” launch places Google’s AI capabilities alongside a cultural and educational debate about representation. It also sets up a benchmark for how generative models may be used in museums, where credibility, provenance, and transparency are likely to be key questions for curators and audiences alike.