AI Data Licensing Glossary: Terms Every Seller Should Know
By Andrew Smart, founder of MON$TER · About
Short, plain answers to the terms you will hear when you sell or license data to AI companies.
- Data licensing
- Granting a buyer the right to use your data under set terms while you keep ownership.
- AI training data
- Text, images, video, audio or records used to teach AI models.
- Recurring license
- A license that renews and pays again, typically $300K–$3M/yr for businesses.
- One-time sale
- Data sold once, often exclusively, typically $50K–$1M.
- Exclusive license
- Only one buyer may use the data. Pays more once, but blocks other buyers.
- Non-exclusive license
- Several buyers may license the same data, which is how recurring income stacks.
- Anonymization
- Removing anything that identifies a person so data can be licensed safely.
- PII
- Personally identifiable information such as names, emails, phone numbers and addresses.
- Data exhaust
- Data a business creates as a side effect of operating: logs, transactions, calls, sensor readings.
- First-person footage
- Video shot from your own point of view, used to train AI and robotics models.
- Approved hour
- One hour of footage a buyer accepts as meeting its brief. Only approved hours are paid.
- Data marketplace
- A platform where data sellers and buyers meet. See our explainer.
- Data broker
- A company that collects and resells personal data, often without consent. Different from licensing.
- The split of what the buyer pays. MON$TER individuals keep 70%.
- Synthetic data
- AI-generated data made to imitate real data. Buyers still pay for real-world data to ground it.
- Data provenance
- A record of where data came from and that you have the right to license it.
- Opt-in consent
- Clear permission from the people in or behind the data before it is used.
- Data room
- A secure space where a buyer reviews a sample before a deal.
Next: how MON$TER works.