- The study investigates privacy risks for domestic workers (DWs) due to AI-driven smart home devices in both employer and personal households.
- Interviews with 18 UK-based DWs and threat modeling using Communication Privacy Management (CPM) reveal intensified surveillance in employer homes through AI features and opaque agency arrangements.
- In their own homes, DWs had more control but faced challenges like gendered administrative roles, unclear AI functionalities, and data retention uncertainties.
- The research proposes a sociotechnical threat model that identifies DW agencies as institutional adversaries and maps privacy risks across interconnected households.
- It outlines social and practical implications to strengthen DW privacy and agency, extending threat models beyond abstract adversaries and single-household contexts.