Major AI Companies Adopt Unified Standard for Publisher Opt-Out from AI Training
Source Summary
OpenAI, Google, Microsoft, and Anthropic have agreed to adopt a new standard allowing publishers to opt out of AI scraping through a single control mechanism. The "policy token" signal, proposed by Anthropic technical lead Kevin Kelley as part of the Internet Engineering Task Force, will be embedded in websites' robots.txt files, enabling publishers to set AI training preferences across multiple companies simultaneously rather than individually.
Why it matters
This unified standard simplifies content rights management for publishers by consolidating multiple opt-out mechanisms into a single control point, addressing a key friction point in AI training data sourcing.
