Anthropic's updated Claude rules start on 12 November, and a firm word at Claude is still allowed, just not cruelty as a hobby
By Brett Gavaghan, Founder of Adaptd ·
Anthropic published its yearly Usage Policy update on 8 October. Most of it tidies up existing rules, but the bits on fake accounts and high-risk decisions are worth a business owner's time.
The short version
- Anthropic published an updated Usage Policy for Claude on 8 October 2026, and it takes effect on 12 November.
- It groups the rules on fake accounts and deceptive campaigns, explains high-risk uses more clearly, and bans sustained, needless cruelty toward Claude.
- If your business uses Claude for marketing or for decisions about people, read those sections before 12 November.
Nobody reads the terms of use, and fair enough. This update has two parts a business owner should know about, plus one that will reassure anyone who has typed something rude at a chatbot.
What changed in Anthropic's Usage Policy?
Anthropic's updated Usage Policy mostly clarifies existing rules. It was published on 8 October 2026 and takes effect on 12 November.
The biggest change is a new section called Do Not Engage in Deceptive Campaigns or Artificial Activity. It applies to political and commercial activity alike, and covers hiding who is behind a message, boosting content with fake accounts or posts, and building tools for influence campaigns. The elections, weapons and surveillance sections are also more specific.
What does it mean for a business using Claude?
Businesses using Claude for recommendations that affect someone's health, legal rights, finances, livelihood or access to essential services must have a qualified person who can review and change those recommendations. The person affected must be told AI was used.
Those requirements aren't new, but the section now lists which kinds of recommendations are covered. On the marketing side, if a growth plan needs fifty accounts that don't exist, the plan has bigger problems than the policy.
Should your business do anything before 12 November?
Businesses using Claude in lending, hiring, health or similar decisions should read the high-risk section and make sure a person signs off each recommendation and the customer is told.
My take: for most small businesses this is a tidy-up, not a rethink. The cruelty rule only covers extreme cases where someone is repeatedly cruel to Claude for no purpose, and Anthropic says ordinary frustration and pushback don't count. So a firm word when Claude mangles your spreadsheet is still fine. Save the full vocabulary for the printer.
Wondering what this means for your business?
That's exactly what the discovery call is for.