Critics link AI training to decades of digital copyright theft

Critics are characterizing current generative AI training practices as an extension of longstanding internet-era copyright infringement rather than a novel technological aberration. A viral Bluesky post by user chalk-idiom argues that software developers have systematically appropriated protected music, images, and writing since the internet’s inception, describing coding culture as inherently lawless. This perspective directly counters industry narratives that distinguish large-scale model training from traditional piracy by emphasizing technical transformation or fair use. The argument suggests that public panic over AI stems from recognizing a historical pattern of unauthorized digital appropriation by technologists. While no specific lawsuit or company was named in the post, this framing aligns with arguments made by plaintiffs in ongoing litigation against major AI firms. Legal experts note that establishing such historical continuity could influence judicial perceptions of intent and industry custom in upcoming copyright rulings regarding training data.