For 500 years, the two surviving paintings of Anne Boleyn have been treated as the closest thing we have to a real image of Henry VIII’s doomed second wife. One is a tiny, damaged portrait now in the National Portrait Gallery in London, the other a larger, more detailed work owned by the Earl of Dalhousie. Both show a dark-haired woman with pale skin, a long neck, and a serious expression—stereotypical features of how we imagine Anne. But new research using facial recognition software says neither painting is likely to be her.

How AI redrew the royal portrait

A team at the University of Bradford ran the two portraits through facial recognition software designed to match faces across different images. The software compared them to contemporary descriptions of Anne, like the account from imperial ambassador Eustace Chapuys, who called her tall, dark-haired, and not conventionally pretty. The AI couldn’t find a reliable match. Instead, it suggested the portraits more closely resemble women who lived around the same time but weren’t Anne—possibly court attendants or relatives of the artists.

The researchers argue these paintings were likely produced years after Anne’s execution in 1536, when her memory had already been tarnished by propaganda. The darker, more severe style fits the period’s fashion for moralizing portraits, not realistic ones. It’s also possible the artists never met Anne and relied on secondhand descriptions or other paintings.

Why these portraits became ‘Anne’

The National Portrait Gallery’s tiny painting, known as the ‘NPG portrait,’ was first recorded in 1857. By then, Anne had been dead for over 300 years, and her reputation had swung wildly. Victorians loved a tragic heroine, so they latched onto this image as the definitive Anne. The larger Dalhousie portrait, painted around 1534, was long assumed to be a near-contemporary likeness because it shows Anne wearing the same necklace she owned in real life. But the AI analysis suggests even that detail might be a later addition to boost the painting’s authenticity.

Art historians aren’t shocked by the findings. Many have already pointed out inconsistencies, like the way Anne’s features change subtly between the two paintings. The NPG portrait shows her with a slightly longer nose and fuller lips than the Dalhousie version. Those small differences wouldn’t matter much if the paintings were meant to be realistic, but in the 16th century, portraits often idealized subjects. Anne’s enemies wanted her remembered as a scheming seductress, while her supporters tried to soften her image. The AI just confirmed what some experts have suspected for decades: these weren’t meant to be portraits at all.

So what did Anne Boleyn really look like?

No one really knows. Contemporary accounts describe her as striking rather than conventionally beautiful, with dark hair, dark eyes, and a sharp wit. Chapuys wrote she was ‘not one of the handsomest women in the world,’ but he admitted she had a ‘certain charm.’ Other descriptions call her tall and graceful, with a long neck—traits that don’t appear consistently in either surviving painting.

The lack of a verified likeness hasn’t stopped artists from reimagining Anne over the centuries. Movies, novels, and even museum displays have turned to these two paintings as visual shorthand, reinforcing the idea they’re accurate. But the AI results add to a growing body of evidence that we’ve been looking at the wrong face all along. It also raises questions about how much we can trust any portrait from this era, when artists often prioritized symbolism over realism.

What happens next?

The University of Bradford team plans to publish their full findings in a peer-reviewed journal later this year. They’re also testing the AI on other disputed royal portraits, including those of Elizabeth I and Mary, Queen of Scots. Meanwhile, the National Portrait Gallery has no plans to pull the paintings from display, but it’s added more context to the label explaining the uncertainty around their subject.

For now, Anne Boleyn’s real face remains a mystery. The AI’s findings won’t change history, but they might finally force us to stop pretending these two paintings tell us what she truly looked like.

What You Need to Know

  • Source: BBC News
  • Published: May 01, 2026 at 14:44 UTC
  • Category: Environment
  • Topics: #bbc · #environment · #climate · #anne-boleyn · #anne-boleyn-portraits · #anne-boleyn-ai-analysis

Read the Full Story

This is a curated summary. For the complete article, original data, quotes and full analysis:

Read the full story on BBC News →

All reporting rights belong to the respective author(s) at BBC News. GlobalBR News summarizes publicly available content to help readers discover the most relevant global news.

Curated by GlobalBR News · May 01, 2026

🇧🇷 Resumo em Português

Pinturas históricas de uma das mulheres mais intrigantes da Inglaterra podem não ser o que sempre se acreditou. Um estudo recente com inteligência artificial jogou luz sobre os famosos retratos que supostamente representam Ana Bolena, segunda esposa de Henrique VIII, levantando dúvidas sobre sua autenticidade. A ferramenta analisou traços faciais e proporções dos quadros, sugerindo que, na verdade, eles poderiam ser de outras mulheres da corte ou até mesmo de nobres desconhecidas.

O caso atinge o Brasil não apenas pelo fascínio histórico que envolve a figura de Ana Bolena — executada sob acusação de traição e adultério em 1536 —, mas também porque expõe como a tecnologia está reescrevendo a história da arte e da realeza europeia. Para pesquisadores brasileiros, a descoberta reforça a necessidade de revisitar fontes iconográficas com ferramentas modernas, algo cada vez mais comum em universidades e museus do país. Além disso, o episódio coloca em xeque a confiabilidade de retratos antigos que, até então, eram tidos como referenciais para entender a aparência de figuras históricas. No Brasil, onde a preservação de acervos artísticos enfrenta desafios, a discussão ganha contornos ainda mais relevantes, especialmente em um momento em que instituições lutam para digitalizar e analisar coleções com auxílio de IA.

Enquanto especialistas debatem a origem dessas imagens, a polêmica abre caminho para novos estudos que podem, inclusive, redefinir a iconografia de Ana Bolena — e, por tabela, de toda a história Tudor.