In an intriguing development in art history, an artificial intelligence (AI) neural network has identified a surprising detail within Raphael's Madonna della Rosa: the face of St. Joseph may not have been painted by Raphael himself.

The painting, housed in Spain's Museo Nacional del Prado, has long been a subject of scholarly debate. Many experts have questioned whether it is an authentic Raphael piece or the collaborative work of the Italian master's apprentices. Recent advancements in AI, however, have provided new insights by analyzing artistic nuances undetectable to the human eye, reinforcing the theory that Raphael may not be the sole artist behind the work.

The study, conducted by a research team from the UK and the US, involved developing an AI algorithm capable of detecting Raphael's distinctive brushstrokes, color palette, and shading techniques. The algorithm was trained using confirmed Raphael works, employing deep feature analysis and a modified version of Microsoft’s ResNet50 architecture. Paired with a Support Vector Machine, this algorithm achieved a remarkable 98 percent accuracy rate in identifying Raphael’s style.

When applied to Madonna della Rosa, the AI delivered inconclusive results on the painting as a whole, but closer examination revealed that while the faces of the Madonna, Child, and St. John bore Raphael's hallmarks, St. Joseph's did not. The AI results align with prior theories that the rendering of St. Joseph’s face is of a noticeably different quality than the rest. While speculation exists that Giulio Romano, one of Raphael’s pupils, may have painted the face, this remains unconfirmed.

"This process illustrates how AI can discern intricate details in artwork that escape even the most trained human eyes," explained Hassan Ugail, a mathematician and computer scientist at the University of Bradford. "Using AI doesn’t replace traditional authentication methods; rather, it serves as a powerful complementary tool."

The Madonna della Rosa, dated between 1518 and 1520, first attracted suspicion in the mid-1800s regarding the authenticity of certain elements. With AI’s confirmation of long-held suspicions, this study exemplifies how modern technology can unlock mysteries in classical art.