One of the most urgent issues confronting writers and other creators right now is the use of copyrighted material for generative AI training.
The large language models that power chatbots like OpenAI’s ChatGPT and Anthropic’s Claude require “training” via the ingestion of vast amounts of text, images, and other materials scraped from the internet or incorporated into databases created by AI companies.
Much of this material is protected by copyright. For the most part, AI companies have not sought permission from rights holders for exploiting their work in this way (nor did creators even start discovering the extent of the companies’ use of their content until a few years ago). They claim permission isn’t needed because AI training falls under the definition of fair use—a limited and transformative use of the material that does not require the copyright owner’s agreement (or remuneration).






