Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Models

Shuning Zhang, Gengrui Zhang, Yibo Meng, Ziyi Zhang, Hantao Zhao, Xin Yi, Hewu Li · 2025

Read Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Models by Shuning Zhang, Gengrui Zhang, Yibo Meng, Ziyi Zhang, Hantao Zhao, Xin Yi, Hewu Li (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-07658.txt.

Description

Zhang and co-authors present a 2025 interview study on how users feel about AI guessing private traits from their social videos. Visual language models can now infer place, background, and income hints from clips. The team ran deep interviews with seventeen users about that power. Most believed the models could indeed read such traits, often with unsettling accuracy.

Fears centered on unwanted identification, misuse of personal data, and constant watching. Wrong guesses worried users too, since errors can still harm. People described shields they try, from cropping to posting less, but doubted those shields hold. They asked platforms and regulators for openness, control, and safeguards built in from the start.

Catalog record

CREATORShuning Zhang, Gengrui Zhang, Yibo Meng, Ziyi Zhang, Hantao Zhao, Xin Yi, Hewu Li
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE67 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-07658

Browse the full library · About the Tower of Babel