AI Art is Theft: Labour, Extraction, and Exploitation, Or, On the Dangers of Stochastic Pollocks

TL;DR

The paper argues that AI-generated imagery, especially from diffusion models, constitutes labour theft by mass-scraping of artists' works without consent, framing the issue through a Lockean labor-right lens and Rawlsian distributive justice. It differentiates among theft senses—heist, plagiarism, and labour theft—and foregrounds labour theft as the central ethical concern due to the nonconsensual extraction of creative labor at scale, the disruption of opportunities, and data-colonial dynamics. By examining consent norms, information-sharing practices, and the speed and scale of AI training, the work contends that AI training on public artworks is unjust and.requires new governance and regulation beyond current copyright norms. The argument has broad implications for all data-driven AI systems and calls for mechanisms that respect creators’ autonomy and fair distribution of benefits. Overall, the paper provides a principled ethical critique of generative AI’s reliance on vast, scraped datasets and outlines pathways for more equitable data practices in AI research and deployment.

Abstract

Since the launch of applications such as DALL-E, Midjourney, and Stable Diffusion, generative artificial intelligence has been controversial as a tool for creating artwork. While some have presented longtermist worries about these technologies as harbingers of fully automated futures to come, more pressing is the impact of generative AI on creative labour in the present. Already, business leaders have begun replacing human artistic labour with AI-generated images. In response, the artistic community has launched a protest movement, which argues that AI image generation is a kind of theft. This paper analyzes, substantiates, and critiques these arguments, concluding that AI image generators involve an unethical kind of labour theft. If correct, many other AI applications also rely upon theft.