Rules for Radical AI: A Counter-Framework for Algorithmic Contestation

AI ethics frameworks—grounded in principles of transparency, fairness, and accountability—have become the dominant mechanism for governing algorithmic systems. However, by treating AI as a neutral technology amenable to technical fixes, they obscure its status as a political project embedded in power, inequality, and historical injustice. In this paper, we introduce the concept of algorithmic contestation—the use of collective, tactical interventions to challenge and reshape the socio-political forces that drive the design and deployment of AI. Drawing on Saul Alinsky’s Rules for Radicals, we derive ten “Rules for Radical AI” that activists and researchers can deploy, including “Reveal the Code,” “Organize a Data Strike,” and “Seize Policy Windows.” We first map mainstream ethics principles onto Alinsky’s tactical logic, then illustrate these dynamics through two case studies—the San Francisco facial recognition ban and a New York City gig worker data strike. Finally, we discuss methodological limits, ethical risks, and future research avenues and issue a manifesto: proper AI accountability arises not from compliance alone but from collective contestation of algorithmic power.