The recent arrival of generative AI models has dramatically impacted a number of creative industries, from independent artists, to musicians, graphics designers, voice actors, and choreographers. The ability to train on artistic creations, and then generate samples of "AI art" has raised alarm, as well as deep questions of copyright, ownership, and fair use. While some of these questions will be resolved by ongoing legal disputes, artists face an immediate threat from more questionable uses of these models to "mimic" their individual artistic styles. In this talk, Professor Ben Zhao will describe some of his own experiences learning about the impact of generative models on the art industry, and how his group joined the fight against AI art mimicry. He will describe the technology underlying diffusion models such as Stable Diffusion, Midjourney and Dall-E2, and present their new research on machine learning tools (Glaze) to prevent AI mimicry. Working closely with independent and commercial artists, Professor Zhao and his team designed, evaluated and released Glaze, a tool that disrupts diffusion models' ability to learn artistic styles of individual artists. Finally, he will discuss other perspectives on generative AI art, and speculate on the future development of generative AI tools.