ABSTRACT
Artificial intelligence (AI) is widely promoted as a neutral and efficient solution to organizational bias, yet its deployment often intensifies gendered and intersectional inequalities. Existing literatures provide partial accounts: algorithmic management emphasizes control without addressing gender; bias research treats discrimination as a technical anomaly; and feminist critiques diagnose patriarchy but lack a mechanism-based explanation of how AI embeds inequality. To address this gap, we theorize feminist algorithmic rationality, a regime in which AI reconfigures managerial logics upstream through three mechanisms: the protocolization of gendered work, the quantification of subjectivity, and corpus feedback loops that normalize exclusion as organizational common sense. By situating this framework historically alongside Taylorism and platform governance, we highlight both continuities in masculinized rationalities and ruptures in the bureaucratization of language and data. Building on these insights, we propose a FeministAI governance framework grounded in intersectionality, transparency, and care ethics, moving beyond technical fixes to reimagine organizational rationality. We argue that, without feminist intervention, AI-enabled decision systems risk institutionalizing inequality by embedding gendered assumptions into organizational categories, metrics, and routines, which we conceptualize as algorithmic patriarchy, where inequality becomes an upstream condition of evaluation rather than a downstream error to be corrected.