Developers adopt code scrambling tactics to block AI scrapers

Software developers are increasingly deploying technical countermeasures, including intentional code obfuscation, to prevent large language models from scraping their work. A representative post on Bluesky illustrates this growing trend, where creators admit to publishing deliberately broken code to confuse automated crawlers. This strategy aims to protect intellectual property and prevent idea theft by degrading the quality of training data harvested without consent. While some developers rely on standard blocking tools, others view active sabotage as a necessary defense against aggressive scraping practices. This phenomenon highlights escalating tensions between the open-source community and AI companies regarding data usage rights. The shift toward adversarial publishing risks fragmenting public technical knowledge bases as individual protection supersedes collective access. Industry observers note that such tactics represent a significant escalation in the ongoing dispute over generative AI training methodologies and copyright enforcement.