Cara, an artists’ platform that filters AI-generated images and offers anti-scraping protections, was hit by three major data scrapes in August. The first scraper, who copied nearly all publicly available artwork on the site, later expressed regret and agreed to help Cara founder Jingna Zhang develop an open-source protection tool, highlighting both the vulnerability of online portfolios and the limited legal safeguards available to artists.

One scrape created a 12-terabyte archive containing 12 million works for less than $10, while others collected 8.5 million links and metadata, and 123,000 images alongside posts, biographies and personal information. Cara says the attacks increased server costs, frightened users and prompted some to delete their portfolios; Zhang launched a $120,000 legal fundraising appeal that had raised more than $100,000 by Thursday. Hugging Face removed personal metadata from one dataset but said it would not remove artwork URLs because it did not host the images itself.