Should AI companies be required to pay artists and authors for using their work to train models?

Generative AI models like GPT and Midjourney devoured the entire internet to learn how to create. Now, the creators of that content want their cut. Proponents argue that tech giants are profiting from 'data laundering'—stripping copyright from billions of images and texts to build products that undercut the original artists. Opponents warn that enforcing copyright on training data is technically impossible and would essentially ban AI development in Europe, killing the region's economic competitiveness.