Despite 88% of organizations reporting regular use of generative AI (GenAI), only 6% achieve significant value, and 67% remain stuck in pilot phases. A value perception gap exists between executives and frontline workers, suggesting adoption is not merely a technology challenge but a human one. Dominant adoption frameworks, such as the Technology Acceptance Model and the Unified Theory of Acceptance and Use of Technology (UTAUT), cannot explain this variance. This study proposes that professional identity threat and organizational power are the missing explanatory factors. Drawing on Identity Threat, Organizational Power, Self-Determination, and Sensemaking Research, this study develops a dual moderation framework in which AI identity threat negatively affects adoption posture, episodic power (discrete managerial mandates and monitoring) amplifies this relationship, and systemic power (embedded structures of governance clarity, fairness, and developmental support) buffers it.

The study employed a quantitative two-phase temporally separated survey design with working professionals who actively use GenAI tools (N = 79). Phase 1 captured AI identity threat, episodic power, and systemic power; Phase 2, administered a mean of 10 days later, captured AI adoption posture as a three-orientation continuum of commitment, compliance, and resistance. Constructs were validated via exploratory factor analysis (α: .736–.895) and xiv hypotheses tested using a five-model ordinary least squares regression.

AI identity threat demonstrated a strong, robust negative effect on adoption posture (b = −.313, p < .001, β = −.451), which was stable across all five model specifications and was confirmed in group comparisons (Cohen's d = .674). Moderation hypotheses were not supported at the current sample size.

Sub-scale analyses of the three adoption orientations revealed a two-pathway mechanism: Identity threat significantly predicted compliance (b = .548, p < .001) and resistance (b = .370, p = .002) but not commitment (b = −.067, p = .372). Systemic power showed a marginal positive direct effect on commitment (b = .122, p = .065), and the joint power configuration significantly moderated commitment formation (IT × EP × SP: b = .164, p = .003). Identity threat activates behavioral compliance and resistance; organizational context cultivates motivational commitment — two distinct causal pathways that existing adoption models have not distinguished.

Theoretically, this study provides the first empirical test of organizational power as a moderator of AI identity threat in technology adoption, introduces a two-pathway model integrating identity threat, power, and self-determination theories, and proposes that adoption framing — positioning GenAI as a capability augmentation rather than cost reduction — suppresses identity threat activation before moderation is required. Practically, organizations measuring adoption through usage metrics are measuring compliance, not commitment. Systemic investment in governance clarity, fairness, and developmental support cultivates the committed adoption that generates organizational value.