He Scraped All of Their Art for AI. Now He’s Collaborating on a Tool to Help Them (www.wired.com)

In a significant development for the AI/ML community, photographer Jingna Zhang and her app Cara, which has attracted 1.5 million artists concerned about unauthorized use of their work for AI training, recently faced three major scraping incidents. One notable event involved an individual posting a 12-terabyte archive containing nearly all of Cara's publicly available images, prompting ethical debates within AI forums. While the scraping led to distress among artists and raised questions about legal protections, the scraper, initially known as MandarinDawnPoppy994, ultimately expressed remorse and agreed to collaborate with Zhang on an open-source tool called Lantern, aimed at safeguarding artists' content. Lantern allows artists to generate a “one-way fingerprint” for their images, notifying them if their work appears in any publicly available AI datasets. This initiative arises from the reality that complete security against scraping is improbable and highlights a regulatory gap in protecting digital content. As Zhang and Heft work on this project, they hope it not only empowers artists with more control over their work but also raises awareness about the broader implications of content scraping in an increasingly AI-driven landscape.