Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-0710
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-08-09T06:33:36.182290Z
status: published
lens: g_space
sector: entertainment
headline: Mystery sitter in Holbein portrait could be Anne Boleyn, AI analysis finds
dek: They are two small sketches by the Renaissance master Hans Holbein: one has long been considered to be a portrait of Henry VIII’s doomed second wife, Anne Boleyn, and the other is of an unknown woman whose name was lost to time. Now researchers using AI have discovered that the unnamed woman might be the tragic queen after all, while the other figure could in fact be Boleyn’s mother. The works, which belong to the royal collection and are known as the Windsor sketch and the Unidentified Woman respectively, were…
gain_title: Researchers used an AI model to compare the Holbein corpus and cluster similar portraits, indicating the unidentified woman belongs in the Boleyn-Howard group and the Windsor sketch aligns with Elizabeth Howard.
problem_title: (none)
trace_subject: (none)
gain_reading: Researchers used an AI model to compare the Holbein corpus and cluster similar portraits, indicating the unidentified woman belongs in the Boleyn-Howard group and the Windsor sketch aligns with Elizabeth Howard.
gain_evidence: "The unidentified woman sat in the Boleyn-Howard cluster, while the Windsor sketch seemed to be closer to images of Elizabeth Howard, Boleyn's mother."
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers at the University of Bradford applied an AI model that compares old master paintings to the Holbein corpus of more than 80 sketches in the royal collection. The system built a matrix comparing each image to every other and clustered close likenesses, placing an unidentified woman in the Boleyn-Howard group and linking the Windsor sketch to Elizabeth Howard.

The finding matters because it could correct an 18th-century inscription error that shaped centuries of Tudor portrait identification, showing how computational clustering can reopen art-historical debate. Uncertainty remains high since most works lack contemporary documentary verification and the authors present the result as a prompt for reassessment rather than a definitive reattribution.
limitation: Attribution remains uncertain because most Holbein corpus lacks contemporary verification and researchers frame findings as opening debate rather than settled identification.
tag: Evidence-backed gain
key_points: University of Bradford team led by Prof Hassan Ugail built a matrix comparing more than 80 Holbein images to cluster close likenesses. | Independent scholar Karen Davies questioned the Windsor sketch because the sitter was light-skinned with red hair while Boleyn was described as darker complexion. | Royal Collection Trust said the identity of the unnamed sitter had long been debated and welcomed further discussion.
rundown: The analysis focused on two sketches in the royal collection, the Windsor sketch long labeled as Anne Boleyn and an Unidentified Woman. Davies noted inconsistent 1700s inscriptions in the corpus, citing an image labeled as Henry Howard that was actually his father.

Ugail previously used a similar AI model to attribute a painting to Raphael. The team compared every image against every other to create a huge matrix, which placed the unidentified woman in the Boleyn-Howard cluster.

Holbein, born in Augsburg and active in Basel before moving to England, painted Tudor court figures including Thomas More executed in 1535 and Boleyn beheaded in 1536. The Hever 'Rose' portrait of Boleyn was separately examined earlier this year as a possible visual rebuttal to witchcraft claims.
sources:
- journalism | The Guardian | https://www.theguardian.com/artanddesign/2026/may/03/unnamed-holbein-portrait-anne-boleyn-ai-analysis | 2026-05-03
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

3fbd0243e95075801c008e47425b862df755d0ec280fdffd384a6407f2f35160

previous

0000000000000000000000000000000000000000000000000000000000000000