Exploring the Boundaries of Content Moderation in Text-to-Image Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Riccio, Piera, Curto, Georgina, Oliver, Nuria
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912046470135808
author Riccio, Piera
Curto, Georgina
Oliver, Nuria
author_facet Riccio, Piera
Curto, Georgina
Oliver, Nuria
contents This paper analyzes the community safety guidelines of five text-to-image (T2I) generation platforms and audits five T2I models, focusing on prompts related to the representation of humans in areas that might lead to societal stigma. While current research primarily focuses on ensuring safety by restricting the generation of harmful content, our study offers a complementary perspective. We argue that the concept of safety is difficult to define and operationalize, reflected in a discrepancy between the officially published safety guidelines and the actual behavior of the T2I models, and leading at times to over-censorship. Our findings call for more transparency and an inclusive dialogue about the platforms' content moderation practices, bearing in mind their global cultural and social impact.
format Preprint
id arxiv_https___arxiv_org_abs_2409_17155
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Exploring the Boundaries of Content Moderation in Text-to-Image Generation
Riccio, Piera
Curto, Georgina
Oliver, Nuria
Computers and Society
This paper analyzes the community safety guidelines of five text-to-image (T2I) generation platforms and audits five T2I models, focusing on prompts related to the representation of humans in areas that might lead to societal stigma. While current research primarily focuses on ensuring safety by restricting the generation of harmful content, our study offers a complementary perspective. We argue that the concept of safety is difficult to define and operationalize, reflected in a discrepancy between the officially published safety guidelines and the actual behavior of the T2I models, and leading at times to over-censorship. Our findings call for more transparency and an inclusive dialogue about the platforms' content moderation practices, bearing in mind their global cultural and social impact.
title Exploring the Boundaries of Content Moderation in Text-to-Image Generation
topic Computers and Society
url https://arxiv.org/abs/2409.17155