AI Appearance Raters Raise Youth Wellbeing Concerns
Psychologist Dawn Branley-Bell described on August 10 how AI appearance-rating websites and apps analyze faces, assign attractiveness scores, compare users in real time, and publicly declare a winner. Her article says these systems can make attractiveness appear scientifically measurable, while a numerical score may carry an illusion of authority or objective fact.
AI appearance-rating websites and apps are using facial analysis to produce numerical attractiveness scores, with some services comparing users in real time and declaring a higher-scoring participant the winner. In an August 10 article, The Conversation reports that such interactions can be livestreamed, with audience members ridiculing the lower-scoring user in chat.
Dawn Branley-Bell, a chartered psychologist and associate professor of digital health and wellbeing at Northumbria University, writes that these tools present attractiveness as if it can be measured with scientific precision. She argues that attractiveness varies across people and cultures, while AI outputs reflect assumptions and biases embedded in training data and system design.
A score can look more authoritative than an opinion
Branley-Bell writes that a critical comment from another person can be recognized as an opinion, whereas a computer-generated numerical score may carry an illusion of authority or objective fact. Her article places appearance-rating systems within a broader online culture of measuring, ranking, and optimizing appearance. Some content in that culture concerns personal grooming, clothing, or exercise, while other material promotes cosmetic procedures, extreme dieting, or dangerous practices.
A June 15 NewBeauty article by Chicago facial plastic surgeon Steven Dayan similarly challenges the premise that a selfie-based score captures attraction. Dayan reported that an AI scanner gave his face an overall score of 6.9 out of 10, while separately rating his smile, expression, and warmth at 8.6 and his hairline at 5.2. He argues that still-image analysis cannot capture dimensions such as expression, movement, emotion, and presence.
Implications for AI assessment products
The reporting raises a technical and product-design question: a facial-analysis output can be numerically precise without establishing that the underlying construct is scientifically measurable or socially appropriate to score.
For teams working with computer vision, the case also underscores that accuracy metrics alone do not resolve product risk. A model can calculate facial proportions while its score remains a judgment about attractiveness rather than a scientifically precise measure of it.
Key Points
- 1AI appearance raters convert facial analysis into attractiveness scores, creating outputs that can be perceived as objective judgments by young users.
- 2The Conversation reports public comparisons and ridicule, making the risk profile broader than private selfie analysis or cosmetic recommendation tools.
- 3Numerical precision does not establish that attractiveness is a scientifically measurable or socially appropriate construct to score.
Scoring Rationale
The story concerns a growing consumer use of facial-analysis systems where model outputs can influence self-image and social behavior. It is relevant to computer vision and responsible AI practitioners, but no new model, product release, regulation, or independently reported technical evaluation is documented.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
