Abstract: Driven by advancements in hardware devices and software algorithms, technology is undergoing a transformative shift from 'mechanical computation' to 'human-machine integration', profoundly reshaping human- technology relationships. Drawing upon and critiquing Actor-Network Theory (ANT), this paper examines the essence of technology as a non-human actor and its dual role in extension and imitation of human 'perception' abilities. From the 'perception' perspective, it reveals how the integration of technology into social governance networks challenges human agency, thereby catalyzing the emergence of a Perceptual Governance model. This governance paradigm emphasizes human centrality, enriching interpretations of technology's role while advocating for a return to human- centered governance narratives. Its implementation relies on three interconnected mechanisms:a holographic data-driven knowledge generation framework, governance dynamics balancing individual and societal needs, and collaborative processes enabling autonomous optimization and adaptive coordination. Although technology functions as a critical actor on par with humans, it primarily serves as a foundational and mediating force, reweaving spatial boundaries, power structures, and governance logics, thereby profoundly influencing the efficacy and outcomes of social governance. Building on this framework, Perceptual Governance is expected to help society break out of the spatial-boundary dilemma, clarify the debate on human-machine relationships, and move towards a path of secure development.