A graduate student at MIT, Alex Kachkine, has developed an innovative and ethically mindful restoration technique that physically superimposes AI-generated digital restorations onto damaged paintings. Using high-resolution scans and generative AI, Kachkine’s method identifies thousands of damaged regions, reconstructs what the original might have looked like, and prints the retouched imagery in finely aligned layers onto a transparent polymer “mask.” This mask is then affixed to the artwork with removable varnish, seamlessly filling in gaps while preserving the integrity of the piece underneath. The process took just 3.5 hours to restore over 5,600 damage areas—approximately 66 times faster than traditional manual restoration.

What sets this breakthrough apart is its reversibility and archival value. The mask can be safely removed if future conservators decide to re-treat the work, and all digital restoration data is retained for posterity. Experts highlight that the film-based approach offers a non-invasive alternative, especially suitable for lesser-known paintings that often languish unseen due to limited restoration budgets. By accelerating and democratizing art conservation, Kachkine’s development opens the door to revealing countless hidden cultural treasures, marrying technological precision with deep respect for historical authenticity.