Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-2159
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-22T10:42:08.304268Z
status: published
lens: p_space
sector: health
headline: Exploring Racial Bias in Rating Facial Attractiveness Among Artificial Intelligence Platforms
dek: Abstract As artificial intelligence (AI) becomes increasingly integrated into aesthetic medicine, its influence on beauty standards and clinical practice warrants closer evaluation, especially among a diverse population. This study aimed to evaluate whether AI-based facial attractiveness platforms assign differing attractiveness scores across ethnic groups despite standardized image characteristics, with the hypothesis that significant differences would exist between ethnic groups and AI platforms, reflecting po…
gain_title: (none)
problem_title: Five AI facial attractiveness platforms gave AI-generated images of White women the highest average scores and Black women the lowest, suggesting bias that could shape beauty standards in aesthetic medicine.
trace_subject: (none)
gain_reading: (none)
gain_evidence: (none)
problem_reading: Five AI facial attractiveness platforms gave AI-generated images of White women the highest average scores and Black women the lowest, suggesting bias that could shape beauty standards in aesthetic medicine.
problem_evidence: White females received the highest average score (8.18) | some AI platforms may exhibit bias toward White females
quick_read: Researchers built a set of 120 stock AI-generated images of young adult women, 30 each from Black, White, Latino and Asian backgrounds, all in frontal view with neutral expressions. They ran every image through five AI-based facial rating websites and compared the scores across ethnic groups and across platforms.

The gap matters because AI tools are becoming part of aesthetic medicine, where they may influence beauty standards and how clinicians talk with patients seeking cosmetic enhancement. The study is cross-sectional and uses synthetic faces of women only, so it cannot show how these scores affect real patients or clinical decisions.
limitation: The images were AI-generated stock photos of young adult women only, and the authors say only that some platforms may show bias, so the results do not establish effects on real patients or men.
tag: Evidence-backed problem
key_points: White females averaged 8.18, followed by Asian (7.79), Latino (7.72) and Black (6.91). | White versus Black faces differed by 1.27 points (p < 0.001); White versus Latino by 0.46 (p < 0.001); White versus Asian by 0.39 (p = 0.05). | Platforms also differed: Prettyscale gave the highest scores (8.50) and Beautyscoretest the lowest (6.65). | The authors say clinicians should understand these trends when consulting patients seeking cosmetic enhancement based on societal standards.
rundown: The study tested whether AI-based facial attractiveness platforms assign different scores across ethnic groups even when image characteristics are standardized. The authors hypothesized differences that would reflect preferential scoring of Eurocentric facial features.

The five platforms were Prettyscale, Attractivenesstest, Beautyscoretest, Hotchat3000 and Attractivenessscore. Post hoc comparisons found significant between-platform differences, including Prettyscale versus Beautyscoretest (+1.85), versus Attractivenesstest (+0.38) and versus Hotchat3000 (+1.54).

The authors conclude that some platforms may exhibit bias toward White females, framing this as relevant to clinical practice and to patients with diverse backgrounds.
sources:
- peer_reviewed | Facial Plastic Surgery | https://doi.org/10.1055/a-2945-2437 | 2026-10-21
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

296d40db4139843ecda6b53bdaeb869ca2ddfd24e2c12415890a3ea326e21cad

previous

0000000000000000000000000000000000000000000000000000000000000000