Playbook for legal and ops stakeholders

Your main question is usually: "What is the site actually signaling, and what is it not claiming?"

Read first

• AI Usage Policy
• Output constraints
• Anti-plausibility
• Glossary

Your risk model

You care most about:

• whether the language overclaims force;
• whether a policy signal is being misrepresented as hard control;
• whether documentation is being mistaken for runtime proof;
• whether public wording is reversible and auditable.

Do

• distinguish intent from enforcement;
• insist on precise wording around AI usage, archives, and scraping;
• keep published posture proportional to what the system really expresses;
• preserve reversibility and traceability.

Don’t

• present robots.txt as a legal shield;
• overstate protection against scrapers;
• imply that every declared signal is respected by every bot.