Authors Guild challenges publishers' use of AI trained on copyrighted works without author consent or compensation

The Authors Guild, a 12,000-member advocacy group for writers, publicly addressed concerns about publishers using copyrighted books and articles to train AI systems (machine learning models that learn patterns from text) without permission or payment to authors—a practice that has become common as publishers seek AI advantages.

Unlike previous AI copyright disputes that focused on tech companies scraping the internet, this targets the publishing industry itself: the companies that profit from author work are now using that same work to build AI tools, creating a direct conflict of interest between publisher interests and writer income.

For working authors, this matters because AI trained on their writing competes with their livelihood—publishers can now generate book summaries, jacket copy, and marketing text using AI trained on author catalogues, reducing demand for freelance writers and potentially lowering advances and royalties across the industry.