Exploring Contractual Issues in AI Transactions – Publications

It’s no surprise that AI is dominating the news cycle. One would be hard pressed to find another trend that has so quickly woven itself into the fabric of legal documents. We are seeing service agreement riders, diligence requests, and representations and warranties specifically related to AI appear in all manner of transactions. To this end, we are going to see more and more AI-related issues and regulation as time goes by.

While generative AI (GenAI) issues get a lot of attention, companies are also focusing on traditional AI, such as machine learning. In fact, machine learning and GenAI technologies were the top two new areas companies were actively investigating and working with in 2023.

INITIAL CONSIDERATIONS

Changing Regulatory Framework

The proliferation of GenAI products—and the focus by corporations and consumers on them—is rapidly leading to various forms of AI regulation. That is why our current advice focuses on disclosure, both on the contract and the policy side. Disclosure is vital because of this changing regulatory environment. Companies cannot comply with regulations if they do not have a thorough inventory of the AI technologies that they and their vendors are using.

Notifications

From a notifications perspective, key items customers will want to consider are:

1. Consent: A customer may require a vendor’s consent to the use of AI products. Consider to what extent traditional AI is already embedded in services and products deployed to a customer, and whether express consent is required in these cases.
2. General disclosure obligation: A vendor may be contractually obligated to keep a customer generally informed of AI usage as part of services and/or products.
3. Information requests: A customer may seek to receive specific information on the use of AI as part of the services provided. A vendor might counterbalance such a right with protecting its commercially sensitive business information and the confidential information of other customers whose data sets are used to train the AI tool or that benefit from the AI tool.
4. Issue notification: If a party detects issues with the use or output of AI, then each party will likely seek a mutual obligation to be promptly notified. Key points of negotiation may include:
   1. The scope of issues
   2. Time period and scope of notification
   3. Consequences of any issues

Key Contract Provisions

A lot of these negotiation concepts are not new, but AI is causing us to think about them a little differently than we would in the traditional context. One focus throughout all of this is how do we think about the traditional provisions that are in the agreement in a way that is informed by this new world of AI-driven technologies.

Testing

Customers will want to build into contracts adequate time to really test the AI before it is deployed. AI products require extensive testing to determine if they are functioning as intended. Consider pilots and soft launches to understand whether a product is ready for production.

Compliance

The contractual allocation of compliance responsibility within the AI ecosystem is becoming increasingly important. Customers will want to make sure that the compliance with law representation is broad enough to cover not just the laws that may be applicable to a vendor but also those applicable to the intended usage or the usage in accordance with the agreement.

IP Ownership/Licensing

The notion of ownership and licensing is unique in an AI discussion. When making this determination, customers will want to consider: Is this AI going to be used and trained on their data? It is one of the driving features as to how to determine which provisions to include.

IP Ownership/Infringement

Customers will want to consider if their IP ownership provisions are broad enough to capture the likely outcomes of the AI product. A major issue is going to be to what extent vendors are willing to stand behind third-party IP protection for the outcomes or new data generated by the AI product.

AI Use Policy

There are two things to keep in mind regarding AI use policies:

1. Requirement for disclosure of use, such that ensuring that internal use and external use are properly disclosed, and the risk of use properly evaluated
2. Any work product requires independent human verification

A NEW WAY OF DOING BUSINESS

AI is not just making news. It is changing how companies do business and will even more so in the future. AI products are increasingly coming to market, and in a widening array of use cases. Contract terms in most transactional areas are being specifically developed to address the issues that crop up. Directly related regulation is also being created, which will continue to make this an evolving area for the foreseeable future.