Generative AI in Art and Literature Creation: Ethical and Creative Dimensions

Abstract

Generative AI, such as GPT-4 and DALL·E 3, is revolutionizing art and literature by enabling novel forms of creation, yet it raises profound ethical and creative challenges regarding authorship, authenticity, and cultural representation. This study proposes a Creativity-Authenticity Framework (CAF) to evaluate AI-generated works across visual art, poetry, and narrative fiction, integrating computational metrics (novelty, coherence) and human expert assessments (emotional resonance, cultural relevance). Analyzing 750 AI-generated outputs from diverse cultural contexts (USA, Japan, Nigeria, Germany), we find that AI achieves high stylistic coherence (82% accuracy) but struggles with emotional depth (48% human parity). Poetry outperforms fiction (68% vs. 35% critic approval), while visual art varies by cultural prompt (e.g., 75% accuracy for ukiyo-e vs. 62% for Afrofuturism). Ethical concerns, including plagiarism risks (18% overlap with copyrighted works) and cultural homogenization, are explored through case studies. We advocate for hybrid human-AI collaboration models, supported by transparent governance, to balance creativity and ethical integrity.