The rapid evolution of artificial intelligence has revolutionized digital art creation, enabling the development of novel methodologies that integrate artistic synthesis with robust intellectual property protection. In this study, we propose an integrated framework that combines Generative Adversarial Networks (GANs), fractal analysis, and wavelet-based turbulence modeling to generate abstract artworks inspired by Jackson Pollock's drip paintings. Beyond emulating Pollock’s dynamic style via neural style transfer, our approach quantitatively characterizes the artworks' intrinsic complexity using fractal dimension and turbulence power spectrum metrics. Importantly, we introduce a comprehensive watermark robustness testing protocol that embeds imperceptible digital watermarks into the generated images and rigorously assesses their resilience against common perturbations—including Gaussian noise, JPEG compression, and spatial distortions. By merging these watermarks with NFT metadata, our framework ensures secure provenance and immutability of digital assets. Experimental results demonstrate the feasibility and efficacy of this multifaceted approach in advancing both artistic innovation and reliable digital copyright protection.