This chapter develops a sociotechnical systems (STS) approach to artificial intelligence (AI). Extending principles from human-centered design (HCD), human factors, and human-centered AI (HCAI), this chapter articulates design practices capable of addressing the complexity, opacity, and scale of modern multilevel AI-STS. Unlike traditional technical models, sociotechnical approaches recognize that innovation emerges from the dynamic interaction of technical, social, and environmental subsystems, each shaped by cognitive, organizational, and cultural processes. The integration of AI into STS introduces distinctive uncertainties, including algorithmic opacity, data biases, shifting organizational structures, and unpredictable multilevel dynamics that produce unintended consequences. These uncertainties complicate both design and governance, requiring adaptive, participatory, and iterative processes that foreground stakeholder diversity, distributed cognition, and accountability. By situating AI within dynamic STS, this chapter emphasizes how micro-level psychological processes, meso-level group and institutional processes, and macro-level sociopolitical and cultural externalities shape technological adoption and adaptation, perceptions of trust, and varying transformative human–machine configurations. In doing so, it further positions the STS framework as essential for addressing the needs of HCAI under conditions of complexity that generate deep uncertainties.