Anthropic's usage policy changes on 12 November: your sensitive AI mu…
On 12 November, Anthropic's usage policy requires a qualified human able to modify, and an informed person. Eleven domains, including employment and grading. The three criterion compliance grid.
You wired Claude into a feature that recommends something to someone: a file, an answer, a note. Since 8 October, you know its terms of use change on 12 November. What the new version asks of your validation screen fits in one sentence, and none of your stories probably writes it down. What you walk away with: a compliance grid per AI feature, and three acceptance criteria to paste into your stories. What changes (PM, PO, designer, design system lead): a human who can modify, an informed person, a record of who validated. Why this week: Anthropic published its usage policy update on 8 October 2026, applicable on 12 November. · · · <!-- order:freeform --> The clause, read on the page and not in a summary Claude's recommendations in a sensitive domain fall under what Anthropic's usage…