Recent research using artificial intelligence led by scientists from the University of Bradford in England suggests that a 500-year-old Raphael painting, the "Madonna della Rosa" ("Madonna of the Rose"), may have been partly painted by another artist, specifically the face of Joseph in the painting.
Raphael, a renowned Renaissance master who lived from 1483 to 1520, is best known for iconic works like "The School of Athens" and "Three Graces."
According to Hassan Ugail, director of the Centre for Visual Computing and Intelligent Systems at the University of Bradford, a new algorithm has been developed to identify authentic Raphael paintings with an impressive 98% accuracy, as reported by CNN on Thursday.
Ugail explained that the authenticity of a painting, such as "Madonna della Rosa" in the Museo del Prado in Madrid, is determined by analyzing 4,000 parameters, including brush strokes, color palette, and hue. Study co-author Howell Edwards, emeritus professor of molecular spectroscopy at the University of Bradford, mentioned that the painting has been the subject of debate for a long time.
According to Edwards, some experts believe that the quality of the composition and painting for the Madonna, Child, and St John far surpasses that of St Joseph, who they believe was added as an afterthought in the workshop.
"Analysis by our AI program has conclusively shown that while the three figures of the Madonna, Christ Child, and St John the Baptist were unequivocally painted by Raphael, the figure of St Joseph was not and was painted by another artist," Edwards added.
Debate around AI in art
Building on prior research, the algorithm concluded that the "de Brécy Tondo" painting is likely an authentic Raphael piece, contradicting previous claims by some art historians who believed it to be a 19th-century copy.
Some art historians criticized the findings, with Ugail stating that they were "a bit dismissive" of the research. He expressed surprise at their reaction, noting that the algorithm analyzed details that are imperceptible to the human eye.
"It will be very, very difficult, no matter how good somebody is, to actually go into that level of detail and create something like that," Ugail told CNN.
Hassan Ugail pictured with the de Brécy Tondo.
University of Bradford
The discussion surrounding the "de Brécy Tondo" contributes to larger conversations about the use of AI in authenticating art, which Ugail believes complements other methods of analysis, such as investigating a work's provenance.
"This is simply another resource," emphasized Ugail, highlighting that the algorithm can help determine if a mysterious artwork warrants further examination.
"He said this is a fast method for determining if a painting warrants further examination. Next, Ugail intends to create an algorithm that can identify pieces by different artists, a development he believes will introduce scientific methods to the process of authenticating art."
"The potential for this kind of tool is huge," he said.
A peer-reviewed paper on the work was published Thursday in Heritage Science journal.