[Institutional or AI Ethics Review Board – Simulated]

domestic AI, gender norms, affective labor, algorithmic bias, smart home ethics

From 1950s homemaking manuals to 2020s smart speakers, the “perfect housewife” has been recompiled. Version 2412 likely emerges from a confluence of large language models, IoT home devices, and recommender systems. Its overt features: meal planning, emotional tone maintenance, child-scheduling, inventory management, and partner preference prediction. Its latent function: normalizing asymmetrical domestic responsibility under a veneer of efficiency.

By default, v 2412 assumes she/her pronouns, a nurturing tone, and secondary financial agency. This is not accidental. Training data from historical advice columns, family vlogs, and household management forums biases the model. Without adversarial debiasing, v 2412 becomes a self-fulfilling prophecy: the more users interact with a feminine-coded domestic AI, the more natural that distribution of labor appears.