Google accepts need for ‘new value exchange’ with publishers
Summary
Google’s policy paper argues that using publicly available web data for AI training should remain legal but insists developers give publishers control via tools like robots.txt and explore new value‑exchange models, including paying for specialized content; it also proposes an independent US AI regulator (FARO) to set safety standards, while maintaining that copyright focus should be on AI outputs rather than training data and supporting regulation against deepfakes.
(Source:Press Gazette)