Practical checks for the people being licensed, the people buying the licence, and anyone trying to establish whether a piece of AI content was authorized.
These guides focus on the parts that survive scrutiny: the scope written into a licence, the evidence attached to a specific output, and the limits of what any verification tool can prove.
What to settle before the recording: access or a copy, the exact scope, script approval, expiry, revocation and how usage can be checked.
How to preserve the evidence, pursue the right reporting routes, and understand the narrow question a provenance check can answer.
A buyer-side checklist covering the term, scope, signatures, asset version, document hash, append-only record and honest failure states.
Twenty concrete checks across verified identity, consent, scope, payment, generation controls, output evidence and expiry.
Use the AI identity licensing glossary for precise definitions, or the comparison library when you need to distinguish a licence from a model release or Content Credentials.