Identifying Intellectual Property in AI Contracts

As artificial intelligence (AI) continues to revolutionize industries, intellectual property (IP) rights in AI contracts are becoming increasingly important. In this article, we will explore the process of identifying the IP in AI contracts and discuss strategies for protecting the interests of both parties.

Identifying the Intellectual Property

In AI-based solutions, identifying the IP at the outset is crucial. This involves understanding the ownership of machine learning algorithms and their implications on competition. One of the primary challenges in AI contracts is determining who owns the machine learning algorithms and statistical models that comprise the solution.

Ownership of Algorithms and Statistical Models

It is fairly straightforward to understand that the software vendor or open-source project owns the algorithms and statistical models that comprise the AI solution. However, understanding the implications of that ownership on competition can be more complex.

The customer typically owns its data, including customer data, business processes, and trade secrets. The thornier question is who owns the learnings derived from machine learning and what restrictions will be placed on the use of those learnings for the benefit of other customers, including the licensee’s competitors.

Negotiating IP Ownership and License Grant Clauses

AI contracts involving machine learning can be much more difficult to negotiate than traditional software licenses. Lawyers on both sides must take the time to fully understand the solution and ensure that IP ownership and license grant clauses are carefully crafted to protect both parties’ business interests.

As a starting point, it is helpful to redline vendor contracts to clearly identify machine learning as a category of IP and draft the contract to ensure that the licensee owns the rights to the learnings. If the software vendor is uncomfortable with that, a discussion should be had about what portion of the learnings the vendor claims an interest in and the extent to which it expects to make the learnings available to competitors.

By including AI-specific definitions in the license agreement, including a definition of machine learning, parties can negotiate a contract that is fair to both sides and easy to interpret if there is a dispute in the future.

Identifying the IP in AI contracts is a critical step in protecting both parties’ interests. By understanding the ownership of machine learning algorithms and their implications on competition, and carefully negotiating IP ownership and license grant clauses, companies can ensure their IP rights are safeguarded in the rapidly evolving world of AI technology.