In this paper, we argue that similarity-based privacy metrics cannot ensure regulatory compliance of synthetic data. Our analysis and counter-examples show that they do not protect against singling out and linkability and, among other fundamental issues, completely ignore the motivated intruder test.

But in practice Generative AI has made significant progress recently, with the actual identifiability of individuals can be highly applications spanning text, code, image, video, speech, and context-specific as different types of information carry different structured data (Sequoia Capital, 2022). Investor interest has levels of identifiability risks depending on the circumstances. However, whether the ChatGPT (Bloomberg, 2023), which has reached 100M resultant synthetic data constitutes personal or anonymous monthly users (Reuters, 2023). This raises the question, as well. Active legal cases against Generative AI companies what constitutes a sufficient level of anonymization.