In my last post, I gave an overview of data licensing, discussing why it’s an important revenue stream for many companies; important for product development, marketing, and other use cases; and general, high-level concerns to consider. In today’s post, I will go in more detail about what issues to look out for when you’re the data owner and want to commercialize it.
When Data is the Core Asset
In some arrangements, the Data Licensing Agreement covers data sets as the core asset being commercialized. Many businesses license out their data sets for various reasons. For example, some companies amass large amounts of data on other companies and sell them as data sets for marketing, sales lead generation, and even for academic or research purposes.
Common Issue Areas
Here are a few common issue areas when you’re the licensor:
- Specific Datasets: What specific data sets are covered under this agreement? Usually, exhibits will list out all these data sets so there’s no ambiguity.
- Access: How will this data be accessed? How will access be provisioned, and will a server need to be replicated?
- Ownership: As licensor, you want to retain ownership over your data and will need appropriate language covering this. Who owns any derived data, if its creation is permitted?
- License Terms: How will you let your customers use this data? Examples include: access, reproduction, distribution, and potentially creation of derivative works, depending on the use case, fees paid, and other factors. A note about derivative works: sometimes, customers want to create “derived data” -- data that is derived from the data you are providing access to, for their own business purposes. Do you allow that, and if so, for what uses? Other license terms include usage in certain geography or industry areas, and should be tailored to each of your customers’ needs.
- AI Use Cases: Are you licensing out your data so that your customers can use it to train their own AI models, or develop their own AI tools? What’s their specific use case? Will you provide a period for the customer to evaluate all or a subset of your data, to assess commercial value? Will you charge for that evaluation? Are there any other customer concerns to address?
Data licensing in AI training and development is very different from traditional data licensing arrangements, and as such, “off-the-shelf” data license agreements aren’t sufficient to adequately address all areas. If you’re planning to license your data for any AI use cases, you should speak with a commercial attorney well versed in licensing, policies, and governance for AI.
