{"id":556358,"date":"2026-05-05T20:36:40","date_gmt":"2026-05-05T20:36:40","guid":{"rendered":"https:\/\/winklersart.com\/?p=556358"},"modified":"2026-05-05T20:36:40","modified_gmt":"2026-05-05T20:36:40","slug":"did-facial-recognition-find-a-lost-portrait-of-anne-boleyn-scholars-debate-whether-a-i-solved-or-merely-muddled-an-art-history-mystery","status":"publish","type":"post","link":"https:\/\/winklersart.com\/?p=556358","title":{"rendered":"Did Facial Recognition Find a Lost Portrait of Anne Boleyn? Scholars Debate Whether A.I. Solved or Merely Muddled an Art History Mystery"},"content":{"rendered":"
\n

Accused of treason, the second wife of Henry VIII lost her head. Now, some researchers argue that she also lost her face among dozens of potentially mislabeled portraits in a royal art collection<\/h2>\n
\n
\n
\n \"Christian\n <\/div>\n
\n

\n

Christian Thorsberg<\/p>\n

| Daily Correspondent<\/span><\/p>\n

\n\"Anne\n <\/picture>
\n

For centuries, the sketch on the left has been identified as Anne Boleyn, while the identity of the woman on the right has been unknown.
\n Royal Collection Trust<\/span>
\n <\/figcaption><\/figure>\n

In a quest to solve a centuries-old mystery that has captured the curiosity of art historians and Renaissance chroniclers, a team of researchers used an A.I. model to try to identify an unknown figure from Tudor history.<\/p>\n

Their findings, published in March in the journal npj Heritage Science, suggest that a 16th-century sketch housed in Britain\u2019s Royal Collection Trust is none other than Anne Boleyn, the ill-fated queen and second wife of Henry VIII executed at the Tower of London.<\/p>\n

The study also suggests a second sketch supposedly portraying Anne Boleyn actually doesn\u2019t, after all. Instead, the researchers say, that portrait shows her mother, Elizabeth Howard.<\/p>\n

It\u2019s a fresh twist to a dispute among scholars about what the queen looked like.<\/p>\n

\u201cI think now we\u2019ve opened up the question,\u201d Karen Davies, an independent historian and the lead author of the study, tells the Guardian\u2019s Lanre Bakare. \u201cIt\u2019s not like we\u2019re making a claim and that\u2019s the thing settled. I hope that there\u2019s a debate about reassessment more widely.\u201d<\/p>\n

\n

\"Sketches\"

\n

Researchers used other sketches by German artist Hans Holbein to train the A.I. model.<\/p>\n

Royal Collection Trust<\/span><\/p>\n<\/figcaption><\/figure>\n

There is indeed already debate about the claim that a new likeness of Anne Boleyn has been identified.<\/p>\n

\u201cWe don\u2019t have a lifetime painted portrait of her that\u2019s absolutely secure, a wonderful painting that we can use as a reference point,\u201d says Charlotte Bolland, a senior curator for research and 16th-century collections at the National Portrait Gallery, to BBC News\u2019 Harriet Bradshaw.<\/p>\n

\n

\"Anne

\n

Known portraits were used for comparative analysis.<\/p>\n

Royal Collection Trust and National Portrait Gallery, London<\/span><\/p>\n<\/figcaption><\/figure>\n

Roughly 500 years ago, German painter Hans Holbein the Younger, an artist lauded by his contemporaries for his work as a portraitist, made his way to England and worked to immortalize the Tudor court. Today, about 85 of his drawings survive in a collection at Windsor Castle. They remain a key source for historians studying the major players in the violent king\u2019s circle.<\/p>\n

But only about 30 have been confidently identified. Others were labeled in the 1700s, and some modern scholars doubt the accuracy of these inscriptions. These 18th-century labels marked one portrait as Anne Boleyn, while another was known to depict an unidentified woman.<\/p>\n

Written evidence suggests that one of the drawings in the Holbein set is of Anne Boleyn, according to BBC News. But which one?<\/p>\n

\n

\"Family

\n

A visualization of facial similarities across family members in the Tudor court, \u201cvalidating biological relationship detection,\u201d according to the study<\/p>\n

Davies et. al.<\/span><\/p>\n<\/figcaption><\/figure>\n

Davies thought modern technology might offer clarity. After all, researchers recently used artificial intelligence to analyze microscopic brushstrokes that seem to confirm the authorship of an El Greco painting and to potentially solve the mystery of a disputed Vermeer canvas.<\/p>\n

So Davies sought the help of Hassan Ugail, a mathematician and scholar of visual computing at the University of Bradford, who previously used A.I. to assess a painting attributed to Raphael. (That research prompted further debate, too.)<\/p>\n

Davies, Ugail and David Stork, who teaches in the materials science and engineering department at Stanford University, used facial recognition to assess portraits of Anne Boleyn\u2019s family members, including her daughter, Elizabeth I, to hunt for elements of resemblance.<\/p>\n

\u201cThe Holbein drawings functioned as working likenesses, technical blueprints for painted portraits,\u201d Stork tells the Times of London\u2019s Mark Sellman. \u201cThat makes them uniquely suited to biometric analysis, which measures bone structure and proportion rather than hairstyle or costume.\u201d<\/p>\n

\n
\n

Did you know? Gallows humor<\/p>\n

One source of information about Anne Boleyn\u2019s appearance is the anecdote that, before she was beheaded, she joked about having \u201ca little neck.\u201d <\/p>\n<\/p><\/div>\n<\/div>\n

The trio\u2019s interpretation of the new analysis suggests that the sketches should be relabeled, and that the unidentified woman is the real Anne Boleyn. They also argue that this result better aligns with written descriptions of the queen\u2019s appearance, too.<\/p>\n

But art historian Bendor Grosvenor, for one, is not convinced. In social media posts made on May 4, he argues the npj Heritage Science<\/em> research conclusions are based on mere \u201cstatistical noise\u201d and questions \u201chow plausible it is to judge facial similarities between portraits of different people by different artists working in different media.\u201d<\/p>\n