Raphael: “The De Brecy Tondo”
In early 2023, the art world was abuzz with the revelation that the de Brecy Tondo Madonna had been authenticated as a masterpiece by Raphael through an AI system. The announcement from Bradford University in the UK reported a 95% resemblance between the Tondo Madonna and Raphael’s Sistine Madonna at the Gemäldegalerie Alte Meister in Dresden. At Art Recognition, we found these revelations intriguing and decided to conduct our own analysis of the painting. Our AI model returned an 85% probability that the artwork was not created by Raphael.
Operational Principles of AI Systems. The discrepancy between the two AI systems developed at Bradford University and Art Recognition stems from their distinct operational principles. Bradford’s AI focuses on facial recognition. It is trained on datasets of faces, and can produce high similarity scores between images depicting the same individual, regardless of variations in face orientation, lighting, or quality. This program, while proficient in recognizing similar faces, isn’t made for artwork authentication.
In the case of the Madonnas depicted in the de Brecy Tondo and the Sistine Madonna in Dresden, Bradford’s program merely indicated that the faces in the two paintings were very similar. This result is not surprising at all, as artists of that period often sought to depict an idealized standard of beauty. The observation that the two Madonnas resemble each other simply confirms that they share a similar aesthetic. However, the authors of the study extrapolated from this observation to conclude that the de Brecy Tondo is undoubtedly by Raphael. This conclusion lacks scientific grounding and is not supported by the study itself. In essence, the authors fabricated a narrative and presented it to the press, which disseminated it without fully understanding its implications.
At Art Recognition, we employ a different approach, considering various artistic elements like chromatics, object placements and compositional elements. Our training datasets, curated by a team of art historians and AI developers, encompass all authentic artworks and a wide range of negative examples, helping the AI model to distinguish genuine art from counterfeits and providing a probability assessment for authenticity determination.
A New Model. After the failure of their face recognition model, the Bradford group developed a new technology more akin to the approach used by Art Recognition. Once again, their AI system returned the response ‘authentic’. In this case, the difference emerged in the datasets used to train the respective AI systems. While the Bradford group’s AI was trained using 49 images, Art Recognition utilized the full dataset of over 100 images. Moreover, while the negative examples used by the Bradford scientists did not include any imitations of Madonna paintings, at Art Recognition we created a dataset of negative examples that includes fake Madonna paintings, but also imitations of other motifs such as allegories and religious scenes.
This difference in the size and composition of the training datasets underscores the significant impact that data selection has on the outcomes of AI-driven art analysis. For AI systems to perform effectively, it is important that their training datasets be as complete as possible, encompassing all known authentic works by an artist. Any omission of relevant data introduces gaps in the system’s understanding, which can in turn affect the final results.
Following this controversy, we decided to release our full Raphael training dataset. The images and documentation can be downloaded following this link.
The growing influence of AI in the art world demands transparency and accountability. The complexity of AI models necessitates rigorous scrutiny by peers and validation from the scientific community before deployment as a market tool. This ongoing debate sheds light on the factors that influence the use of AI in determining art authenticity and explores how to effectively employ this tool within the art community.
