This study explores the profound implications of AI-generated imagery, focusing on the limits and possibilities of synthetic faces and images within contemporary visual culture. It examines the underlying logic of AI image generation, emphasising the shift from one-to-many to many-to-one processes driven by vast datasets and latent spaces. The discourse highlights critical issues such as the epistemic and legal challenges of originality, authorship, and intellectual property, alongside the emergence of technologies like GANs and deepfakes, which blur the boundaries between real and artificial. The comparison between human and algorithmic perception introduces the concept of “algorithmic similarity,” questioning how machines recognise and cluster images. Furthermore, the paper discusses the socio-cultural significance of authenticity, the remediating influence of AI on older media, and the aesthetic tension between visual accuracy and distortion. Ultimately, it advocates for a nuanced understanding of AI imagery as both a disruptive force and a catalyst for rethinking truth, representation, and creativity in the digital age.