AI tools for Nazi‑looted art restitution

In late July 2026, a group of business‑school professors in France unveiled an artificial‑intelligence system that scans extensive provenance databases to flag artworks potentially stolen by the Nazis. The tool, which matches images, descriptions and ownership histories, was distributed to museums worldwide to aid restitution claims. Around the same time, the Harvard Art Museums piloted a generative‑AI project that animated a 1750 portrait, testing AI‑driven storytelling for visitor engagement.

A week later, researchers at Santa Clara University in California introduced the AI Provenance Assistant, a natural‑language query platform that normalises variant spellings, translations and erroneous entries across museum records, auction catalogs and archives. Targeting the roughly 100,000 missing Nazi‑looted pieces, the system links fragmented clues to accelerate the return of cultural property to heirs, complementing other initiatives such as the Fraunhofer‑run KIKU app.