A mechanical engineering student at the Massachusetts Institute of Technology (MIT) has developed a new AI-based method for physically restoring damaged artworks – a potential breakthrough in art conservation.
Anyone who has ever ventured behind the scenes of a museum will know that, beyond the works on display, there is a vast hidden collection of thousands of paintings and artefacts stored in basements and off-site archives. Many of these are considered too damaged or too costly to restore and therefore remain permanently out of public view.
Thanks to a new method developed by MIT PhD candidate and long-time art aficionado Alex Kachkine, that may be about to change. His technique could enable museums to bring long-forgotten masterpieces out of storage and back into the public eye.
As detailed in the journal Nature, the method involves printing the restoration onto an ultra-thin polymer film – essentially a removable mask – and aligning and affixing it to the original painting. One of the key advantages of this method is that the mask can be easily removed without damaging the artwork, and a digital copy of the restoration can be stored for future reference by conservators.
The innovation lies not just in the speed and accuracy of the process, but also in its transparency. “Because there’s a digital record of what mask was used, in 100 years, the next time someone is working with this, they’ll have an extremely clear understanding of what was done to the painting,” Kachkine explains. “And that’s never really been possible in conservation before.”
A new tool restores damaged paintings in hours using a polymer film mask. Developed by Alex Kachkine, a mechanical engineer at the Massachusetts Institute of Technology, it restored a 15th-century painting in 3.5 hours, digitally filling 5,612 regions with 57,314 colours. pic.twitter.com/ESqmBemgX1
— New Scientist (@newscientist) June 16, 2025
Kachkine demonstrated the technique on a severely damaged 15th-century oil painting. The AI identified 5,612 areas requiring restoration, down to the exact hues and pigment mix. The entire process took just 3.5 hours from start to finish – 66 times faster than traditional manual restoration methods.
The potential applications are significant. This approach could help museums and galleries reduce the costs and complexity of restoring their vast backlogs of damaged works. ‘There is a lot of damaged art in storage that might never be seen,” says Kachkine. “Hopefully with this new method, there’s a chance we’ll see more art, which I would be delighted by.”
Kachkine began this project as a side venture while studying for his PhD in Mechanical Engineering. Drawing on his background in technology and traditional art restoration, he explored how existing AI algorithms, trained on large datasets of historical artworks, could learn stylistic patterns and fill in missing or damaged areas. While digital restorations have typically remained separate from the original canvas, Kachkine’s breakthrough lies in physically applying the result to the artwork itself via a transparent polymer mask.
In fact, the more damaged the painting, the more suitable it is for this approach. Nevertheless, not everyone is convinced. Can an AI-restored version truly reflect the artist’s original style and intent? Kachkine acknowledges these ethical questions but argues that the alternative is worse.
This method could also prevent disastrous results like the infamous ‘Monkey Christ’ fresco restoration in Spain. Yet, as with any new technology in the arts, fundamental questions remain. Should museums embrace these high-tech restorations, or opt for digital displays of reconstructions instead, preserving the original in its authentic, albeit damaged, state?
