A participatory red-teaming project in the Global South created the PLACES dataset of 26k T2I failure examples that reveal unique cultural and linguistic harms missed by existing safety frameworks.
Title resolution pending
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 4
citation-polarity summary
roles
background 4polarities
background 4representative citing papers
Thematic analysis of 43 AI contestation cases, using Bovens's relational accountability model, produces categories of demands from below, institutional pushback, outcomes, and contextual factors shaping effective contestation.
Longitudinal analysis of Reddit posts shows human-AI romance discourse evolving from intimate personal stories to focus on platform governance, technical problems, and societal impacts.
Generative LMs in laissez-faire open-ended prompting settings disproportionately generate subordinated portrayals of minoritized race, gender, and sexual orientation identities at rates hundreds to thousands of times higher than empowering ones.
A new toolkit with cards and maps enables AI designers to juxtapose values and harms in early concept stages, shown valuable in designer surveys and interviews.
Generative AI boosts attackers' ability to create harmful content at scale while also enabling defenders to detect threats, support users, and improve moderation processes.
An online experiment finds that showing users an overview of an AI's values reduces reliance on AI suggestions during writing tasks.
A scoping review of AIES and FAccT literature concludes that AI trustworthiness research prioritizes technical precision over social, ethical, and institutional factors, leaving the sociotechnical nature of AI systems underexplored.
citing papers explorer
-
Going PLACES: Participatory Localized Red Teaming for Text-to-Image Safety in the Global South
A participatory red-teaming project in the Global South created the PLACES dataset of 26k T2I failure examples that reveal unique cultural and linguistic harms missed by existing safety frameworks.
-
Push and Pushback in Contesting AI: Demands for and Resistance to Accountability
Thematic analysis of 43 AI contestation cases, using Bovens's relational accountability model, produces categories of demands from below, institutional pushback, outcomes, and contextual factors shaping effective contestation.
-
Technically Love: The Evolution of Human-AI Romance Discourse on Reddit
Longitudinal analysis of Reddit posts shows human-AI romance discourse evolving from intimate personal stories to focus on platform governance, technical problems, and societal impacts.
-
Laissez-Faire Harms: Algorithmic Biases in Generative Language Models
Generative LMs in laissez-faire open-ended prompting settings disproportionately generate subordinated portrayals of minoritized race, gender, and sexual orientation identities at rates hundreds to thousands of times higher than empowering ones.
-
Developing an AI Concept Envisioning Toolkit to Support Reflective Juxtaposition of Values and Harms
A new toolkit with cards and maps enables AI designers to juxtapose values and harms in early concept stages, shown valuable in designer surveys and interviews.
-
How Generative AI Empowers Attackers and Defenders Across the Trust & Safety Landscape
Generative AI boosts attackers' ability to create harmful content at scale while also enabling defenders to detect threats, support users, and improve moderation processes.
-
Framing an AI with Values Reduces AI Reliance in AI-supported Writing Tasks
An online experiment finds that showing users an overview of an AI's values reduces reliance on AI suggestions during writing tasks.
-
Understanding AI Trustworthiness: A Scoping Review of AIES & FAccT Articles
A scoping review of AIES and FAccT literature concludes that AI trustworthiness research prioritizes technical precision over social, ethical, and institutional factors, leaving the sociotechnical nature of AI systems underexplored.