The Bot Cruelty Plan: Three Decisions for a Support Bot Built on Claude
Template Free
Template The Bot Cruelty Plan
Mathew PaquetTrust & Risk · Template

The Bot Cruelty Plan: Three Decisions for a Support Bot Built on Claude

Decide what your bot does when a customer is cruel to it, how it says it's a bot, and how it answers "Do you have feelings?", before a customer finds out for you.

4 min
5 checks
3 copy-ready templates
6 sources cited

One email gets you a library card. It opens this template and all 53 resources in Mathew's library, now and later.

Mathew's LibraryLibrary card
Free access
[email protected] All 53 resources
Already have a card on another device? Enter the same email. You get it here right now and a copy by email, plus The Operator Notes. Unsubscribe anytime. Privacy
Thanks
Open the template
Your library card opens everything here
I'm the CEO of myMarketing and Ryv AI, and I make a video every day on what new AI changes for marketers and founders. Every resource here is built from those videos and fact-checked.

Swipe through the preview

What's inside

What the new rule covers, and the gap it leavesYour customers are covered. The end-chat button isn't on your bot.
Decision one: your abuse pathA threshold, one line, a handoff, a log. Fill-in block included.
Decision two: say it's a botWhat Anthropic's policy requires, and a disclosure line to copy.
Decision three: "Do you have feelings?"Pick your answer and put it in your bot's instructions.
The test before a customer runs itA tick-box checklist you can run on your own bot today.

Quick win you can do today. Open your support bot today, ask it if it's a person and if it has feelings, and write down the threshold where a cruel chat goes to a human.

You might also want

Trust & Risk
Template The Bot Cruelty Plan
The Bot Cruelty Plan: Three Decisions for a Support Bot Built on ClaudeFree · 4 min