How AI is Restoring Famous Paintings: The Technology Saving Our Art History

What if a painting damaged 350 years ago could look new again? What if a symphony Beethoven never finished could finally be heard? AI is making both of these things happen — and the results are stunning.

Bringing The Night Watch Back to Life

Rembrandt's The Night Watch is one of the most famous paintings in the world. But in 1715, it was cut down to fit a new location — removing whole sections of the original scene. Those cut pieces were lost for centuries.

In 2019, the Rijksmuseum in Amsterdam used AI to change that [Operation Night Watch]. The team fed thousands of high-resolution images of Rembrandt's other works into an AI system, which learned his exact brushstroke style, colour palette, and lighting technique. The AI then reconstructed the missing sections with such precision that visitors can now see the painting as Rembrandt originally intended it.

Catching Forgers

Art forgery is a billion-dollar problem. A fake painting can fool even expert eyes. But AI is making it much harder to cheat. When it analyses a painting, AI looks at microscopic details — the exact pressure of each brushstroke, how paint layers were built up over time, the chemical signature of different pigments. Every great artist has a unique fingerprint in how they work. Studies show AI can detect forgeries with up to 97% accuracy [Elgammal et al., Rutgers].

Finishing Beethoven's 10th Symphony

Ludwig van Beethoven died in 1827, leaving only sketches for his 10th Symphony. For almost 200 years, that was where the story ended. In 2021, a team of musicologists and AI researchers [Karajan Institute / Telekom] fed Beethoven's complete works into an AI system. The AI learned his musical patterns — how he built tension, how he resolved it, his favourite chord progressions — and completed the symphony from his sketches. It was performed live to an audience in Bonn.

Cataloguing Millions of Artworks

The world's museums hold over 700 million objects. Most of them have never been properly described or photographed. Google Arts & Culture uses AI to help museums catalogue their collections at scale — identifying objects, suggesting descriptions, spotting connections between artworks from different countries and centuries.

A Tool, Not a Replacement

AI does not replace art historians or restorers. The human experts still make the final decisions about what is authentic, what to restore, and how. AI is the tool that makes their work faster, more precise, and more powerful than ever before.