Did Facial Recognition Find a Lost Portrait of Anne Boleyn? Scholars Debate Whether A.I. Solved or Merely Muddled an Art History Mystery

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

Christian Thorsberg | Daily Correspondent

May 5, 2026 4:36 p.m.

Share

Copy linkEmailSMSFacebookXRedditLinkedInBlueskyPrint

Add as preferred source

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.

Their findings, published in March in the journal npj Heritage Science, suggest that a 16th-century sketch housed in Britain’s 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.

The study also suggests a second sketch supposedly portraying Anne Boleyn actually doesn’t, after all. Instead, the researchers say, that portrait shows her mother, Elizabeth Howard.

It’s a fresh twist to a dispute among scholars about what the queen looked like.

“I think now we’ve opened up the question,” Karen Davies, an independent historian and the lead author of the study, tells the Guardian’s Lanre Bakare. “It’s not like we’re making a claim and that’s the thing settled. I hope that there’s a debate about reassessment more widely.”

There is indeed already debate about the claim that a new likeness of Anne Boleyn has been identified.

“We don’t have a lifetime painted portrait of her that’s absolutely secure, a wonderful painting that we can use as a reference point,” says Charlotte Bolland, a senior curator for research and 16th-century collections at the National Portrait Gallery, to BBC News’ Harriet Bradshaw.

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’s circle.

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.

Written evidence suggests that one of the drawings in the Holbein set is of Anne Boleyn, according to BBC News. But which one?

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.

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.)

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’s family members, including her daughter, Elizabeth I, to hunt for elements of resemblance.

“The Holbein drawings functioned as working likenesses, technical blueprints for painted portraits,” Stork tells the Times of London’s Mark Sellman. “That makes them uniquely suited to biometric analysis, which measures bone structure and proportion rather than hairstyle or costume.”

The trio’s 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’s appearance, too.

But art historian Bendor Grosvenor, for one, is not convinced. In social media posts made on May 4, he argues the npj Heritage Science research conclusions are based on mere “statistical noise” and questions “how plausible it is to judge facial similarities between portraits of different people by different artists working in different media.”

Christian Thorsberg | Read More

Christian Thorsberg is an environmental writer and photographer from Chicago. His work, which often centers on freshwater issues, climate change and subsistence, has appeared in Circle of Blue, Sierra magazine, Discover magazine and Alaska Sporting Journal.