Generative AI and Creative Self-Efficacy: A Systematic Review

• Systematic review of generative AI and creative self-efficacy (2021–2025) • Effects of generative AI on creative confidence depend on design and context • Human agency and attribution clarity shape AI-related creative self-beliefs Creative self-efficacy shapes motivation, persistence, and engagement in creative work. As generative artificial intelligence systems become increasingly integrated into educational, professional, and artistic contexts, questions have emerged regarding how these technologies influence individuals’ creative self-beliefs. This systematic review synthesizes research published between 2021 and 2025 to examine how generative AI is associated with creative self-efficacy, creative confidence, and related constructs during a period of widespread adoption. Following PRISMA guidelines, we analyzed 42 peer-reviewed studies across educational, professional, and creative domains. Fourteen studies directly examined relationships between the use of generative AI and creative self-beliefs, while additional studies provided contextual insight into how creative self-efficacy was conceptualized and measured during the AI integration period. Findings across the reviewed literature were heterogeneous, varying across domains and study designs. Across studies, recurring patterns indicate that perceived control, attribution clarity, and opportunities for human agency are frequently associated with differences in reported creative confidence and self-efficacy. When generative AI was positioned as a supportive tool within transparent co-creation processes, studies more often reported neutral or positive associations with creative self-beliefs. When AI outputs were perceived as displacing human authorship or obscuring individual contributions, studies more commonly reported uncertainty or reduced confidence. Bandura’s four sources of self-efficacy remain relevant in AI-mediated contexts but appear to operate differently in human–AI collaboration. Overall, the review suggests that generative AI does not uniformly enhance or diminish creative self-efficacy; its psychological implications depend on contextual, pedagogical, and design-related factors.