Are You Comfortable Sharing It?: Leveraging Image Obfuscation Techniques to Enhance Sharing Privacy for Blind and Visually Impaired Users

A research team led by Satabdi Das has published groundbreaking work addressing a critical privacy gap for Blind and Visually Impaired (BVI) individuals. Their CHI 2026 paper, 'Are You Comfortable Sharing It?: Leveraging Image Obfuscation Techniques to Enhance Sharing Privacy for Blind and Visually Impaired Users,' tackles the unique challenge BVI users face when sharing images they cannot visually verify. The researchers identified that accidental sharing of sensitive content—from personal moments to embarrassing shots—poses significant privacy and social risks, as BVI individuals often rely on others to describe images or use imperfect screen readers that might miss contextual details.

The team conducted a comprehensive study with 20 BVI participants, evaluating various image filtering techniques including pixelation, blurring, and content-aware obfuscation across different sensitivity levels and sharing contexts (family, friends, strangers). Key findings revealed pixelation was the least preferred method, while preferences for other filters varied significantly based on both image type and intended audience. Participants reported substantially greater comfort sharing filtered versus unfiltered images across all audiences. Based on these results, the researchers developed concrete design guidelines for implementing AI-powered privacy filters in social media and messaging platforms, suggesting systems should offer multiple obfuscation options with clear, accessible descriptions of what each filter does to the image content.