Legal Scrutiny Mounts Over AI Training on Copyrighted Books

23 Aug 2026, 8:30 pm · 20d ago · 1 min read · TechCrunch AI

The legal battle surrounding the use of copyrighted books and creative works to train large language models remains deeply contentious across the artificial intelligence sector. Generative AI developers routinely scrape published literature without explicit author consent or compensation, arguing that ingestion falls under fair use doctrines. However, authors and creative industry stakeholders contend that unauthorized training undermines their intellectual property and economic livelihoods. As global courts and regulators weigh copyright infringement claims, the resulting legal precedents will fundamentally reshape how generative AI companies source training datasets and manage intellectual property compliance.