The booming use of text-to-image generative models has raised concerns about their high risk of producing copyright-infringing content. While probabilistic copyright protection methods provide a probabilistic guarantee against such infringement, in this paper, we introduce Virtually Assured Amplification Attack (VA3), a novel online attack framework that exposes the vulnerabilities of these protection mechanisms. The proposed framework significantly amplifies the probability of generating infringing content on the sustained interactions with generative models and a lower-bounded success probability of each engagement. Our theoretical and experimental results demonstrate the effectiveness of our approach and highlight the potential risk of implementing probabilistic copyright protection in practical applications of text-to-image generative models. Code is available at https://github.com/South7X/VA3.

翻译：文本到图像生成模型的广泛应用引发了对其产生版权侵权内容高风险性的担忧。尽管概率性版权保护方法能对此类侵权行为提供概率性保证，但本文提出了一种新颖的在线攻击框架——虚拟确保放大攻击（VA3），揭示了这些保护机制的脆弱性。该框架通过持续与生成模型交互，并降低每次交互的成功概率下界，显著放大了生成侵权内容的概率。我们的理论与实验结果表明，该方法具有有效性，并突显了在文本到图像生成模型的实际应用中实施概率性版权保护的潜在风险。代码已开源至 https://github.com/South7X/VA3。