Artists are concerned about generative AI models that can mimic their unique styles. To counter this, tools add small adversarial perturbations to artworks to prevent style mimicry. However, this study shows these tools are ineffective. Low-effort methods like image upscaling can bypass protections, and user studies show existing tools can be easily circumvented. Adversarial perturbations cannot reliably protect artists from style mimicry, and alternative solutions are needed. The study evaluates three popular style protection tools: Glaze, Mist, and Anti-DreamBooth. These tools add perturbations to images to prevent mimicry. However, the study shows that simple methods like Gaussian noise or upscaling can bypass these protections. The results show that robust mimicry methods can produce images indistinguishable from those generated from unprotected art. The study also finds that existing evaluations of these tools are flawed. They fail to account for the capabilities of resourceful forgers and rely on automated metrics that are unreliable for measuring style mimicry. A unified and rigorous evaluation is needed. The study proposes four robust mimicry methods: Gaussian noising, DiffPure, Noisy Upscaling, and IMPRESS++. These methods are tested against the three protection tools. The results show that all protections can be bypassed, leaving artists vulnerable to style mimicry. The study also highlights the limitations of current protection tools. They are not adaptive and cannot be updated once deployed. This makes them ineffective against evolving mimicry methods. The study urges the development of alternative protective solutions. The study concludes that adversarial perturbations cannot reliably protect artists from style mimicry. Existing tools provide a false sense of security and leave artists vulnerable. Alternative solutions are needed to protect artists from the misuse of generative AI.Artists are concerned about generative AI models that can mimic their unique styles. To counter this, tools add small adversarial perturbations to artworks to prevent style mimicry. However, this study shows these tools are ineffective. Low-effort methods like image upscaling can bypass protections, and user studies show existing tools can be easily circumvented. Adversarial perturbations cannot reliably protect artists from style mimicry, and alternative solutions are needed. The study evaluates three popular style protection tools: Glaze, Mist, and Anti-DreamBooth. These tools add perturbations to images to prevent mimicry. However, the study shows that simple methods like Gaussian noise or upscaling can bypass these protections. The results show that robust mimicry methods can produce images indistinguishable from those generated from unprotected art. The study also finds that existing evaluations of these tools are flawed. They fail to account for the capabilities of resourceful forgers and rely on automated metrics that are unreliable for measuring style mimicry. A unified and rigorous evaluation is needed. The study proposes four robust mimicry methods: Gaussian noising, DiffPure, Noisy Upscaling, and IMPRESS++. These methods are tested against the three protection tools. The results show that all protections can be bypassed, leaving artists vulnerable to style mimicry. The study also highlights the limitations of current protection tools. They are not adaptive and cannot be updated once deployed. This makes them ineffective against evolving mimicry methods. The study urges the development of alternative protective solutions. The study concludes that adversarial perturbations cannot reliably protect artists from style mimicry. Existing tools provide a false sense of security and leave artists vulnerable. Alternative solutions are needed to protect artists from the misuse of generative AI.