<![CDATA[Artificial intelligence has fundamentally transformed visual arts through generative systems that support ideation, image generation, and collaborative creative production. The widespread adoption of diffusion models, multimodal systems, and text-to-image technologies has raised critical questions surrounding creativity, authorship, human agency, and ethical responsibility. This study conducted a systematic literature review using the PRISMA framework, drawing on peer-reviewed publications from 2020 to 2025 across major academic databases, analyzed through narrative and thematic synthesis. A clear technological evolution is evident, from rule-based computational art to diffusion-based and multimodal AI systems. While these technologies accelerate ideation and visual exploration, increased productivity does not automatically yield greater originality, conceptual depth, or artistic quality. Professional artists consistently outperform novice users in concept development, prompt construction, visual evaluation, and contextual interpretation. Public and expert assessments of AI-assisted artworks are shaped more by perceived human intention, authenticity, and authorship attribution than by technical image quality alone. Persistent challenges remain around intellectual property, creative responsibility, explainability, and the measurement of distributed creative agency. AI is best understood as an augmentative collaborator rather than an autonomous creator. Meaningful artistic production emerges from sustained interaction between computational generation and human intention, expertise, and ethical stewardship. The study's principal contribution is an integrated conceptual framework linking technological capability, human creative processes, distributed agency, artistic evaluation, and governance, serving as a foundation for future research, professional practice, and policy development in AI-assisted visual arts.]]>
