This paper examines how emerging public experiments in red-teaming generative AI (genAI) systems engender new formations of ‘experimental publics,’ shaped by competing visions of expertise, democratic oversight, and AI safety. Interrogating the practice of genAI red-teaming, the authors analyze distinctions between instrumental and deliberative... See more

Regulatory agencies should be charged with evaluating the most important algorithms, and individuals should have a “right to an explanation” when A.I. systems make decisions that shape their lives. And yet he admits that, even if such reforms are put into place, there will be reasons to doubt “the compatibility of democracy with the structure of... See more