As the pace of technological change continues to accelerate with the widespread adoption of cloud computing and the increased use of AI/machine learning, our clients are more regularly turning to technology collaborations to reach their business goals. Those collaborations may take the form of a traditional joint venture or other joint development arrangements, but collaborative technology development is inherent in many outsourcing matters as well. Whatever the form of technology collaboration, fundamental issues of IP allocation can plague a poorly-crafted deal. Because such collaborations often result in technology that’s critical to the business, careful consideration should be given to how IP rights are provided by each party to the collaboration, how newly created IP rights are allocated, and what happens when the collaboration ends.
Technology collaborations can take many forms, and are common across many industries. Depending on the type of collaboration, such an arrangement may be structured as the formation of a joint venture entity, as a bilateral strategic alliance, or as a joint development agreement. Our clients consider in this regard the likely duration of the collaboration, the scope of the collaboration, and its connection to the core business of each party. Such collaborations can exist between competitors, between complementary players, and between customers and service providers. But it is important for clients to understand that even commercial arrangements that don’t fall within these categories often involve technology collaboration—for example, a provider offering outsourced machine learning services often can’t provide those services without the substantial collaboration of the customer.
Whatever the form of collaboration, there are several common IP issues that arise:
- How is the background intellectual property and data of each party—the IP and data that exists as of the date of the commencement of the collaboration—made available to the joint project?
- How do the parties allocate ownership of the resulting IP/data in a way that protects each party’s core business interests?
- What happens when the collaboration ends (whether amicably or in acrimony)?
We are often called upon both to document such collaborations, and bring to this exercise experience in how such collaboration agreements often fail.
III. Why This Topic Matters Now
This topic matters now more than ever in light of the transformational technological changes that have accelerated over the last several years. In addition to the continued penetration of cloud computing, AI and machine learning applications have quickly moved from the theoretical to the intensely practical. Even the most competent companies often lack the internal expertise to quickly leverage these technologies, and turn to collaboration partners to meet their business goals. In addition, the lingering effects of the COVID-19 pandemic and the current uncertainty as to the trajectory of the economy have forced companies to consider how to bring new technologies to market quickly and with lower risk, and collaborating with a technology partner often fits that bill.
In addition, AI and machine learning have introduced new collaboration paradigms that don’t fit neatly in to previous models of joint development. Rather than a combination involving comparable technologies, collaborations in this area often involve extremely asymmetrical technology contributions, with the customer contributing industry know-how and data, and the service provider providing codes and algorithms. But just as in traditional technology collaboration, the whole is very much more than the sum of its parts, putting tremendous pressure on the agreement provisions allocating the rights to exploit the results of the collaboration.
Whether the collaboration involves AI or not, by joining forces with a technology partner, companies may streamline development pipelines that can be very expensive and time consuming, and therefore market products, processes, or services more rapidly. Simultaneously, however, these arrangements present potential pitfalls within the collaboration, including loss of IP rights. For example, in a trade secrets suit involving a failed collaboration, a court denied a motion to dismiss a breach of contract and theft of trade secrets claim because the suing party sufficiently defined the trade secrets at issue with sufficient particularity in the collaboration.1
IV. Recommendations and Best Practices
IP Diligence. While the concept of “due diligence” comes from the M&A context, preparing for a technology collaboration deal involves many of the same considerations centered around the question of what IP/data the technology partner brings to the table. Our clients seek answers to the following questions: (1) what rights does the technology partner own, and how are they protected; (2) what encumbrances are those rights subject to, such as pre-existing licenses or source code escrow arrangements; and (3) what third-party rights may be implicated by the joint project. In addition to addressing such issues in the representations and warranties, technology collaborations often benefit from the substantive review of these issues. Clients should expect reciprocal diligence on their own contributions.
IP Contributions. Once a client has a handle on what IP the technology partner owns and may contribute to the joint project, the parties need to consider how that technology (and related rights) will be made available for that joint project. In an entity-based joint venture, such rights may be assigned to the venture itself, but, more commonly in bilateral agreements, the technology collaborators each license rights for use in the joint project. Such a license should address the use of that technology, rights to improvements, and protection and enforcement. Special consideration should be given to any ongoing rights that may be needed upon the termination of the arrangement.
Collaboration Scope. Business partners to collaboration deals very often have multiple avenues of potential joint work. In drafting collaboration agreements, parties should avoid “mission creep” by narrowly and specifically defining the purpose and scope of the joint development. Clearly and explicitly defining any IP not being contributed is equally essential to include in the collaboration agreement.
Collaboration IP. Once the diligence is conducted, the contributed IP is defined, and the parameters of the collaboration are documented, the parties must next decide how the jointly developed IP and data (“Joint IP”) resulting from the collaboration should be treated. This includes how Joint IP should be owned, rights to seek legal protection of Joint IP, commercial exploitation rights, rights after the termination of the collaboration, and rights to derivative IP, among others. All such issues should be specifically addressed and defined in the collaboration agreement. The key considerations are set forth below:
- Parties often default to a concept of joint ownership of Joint IP, but we recommend such a model be avoided—joint ownership tends to devalue the resulting IP, complicating efforts to protect, enforce, and monetize such IP. For example, a court dismissed a patent infringement claim brought by one party to a development and commercialization agreement against the other on the basis that the counterparty was a joint owner of the patent, a result that the plaintiff likely did not contemplate when negotiating the deal.2
- Collaborations in which one party is contributing only data are especially complex to document, as the data often acts as a key catalyst to the success of the collaboration, but the technology provider can easily exploit the resulting IP without infringing any rights in the underlying data.
- Whichever party owns the Joint IP, rights to exploit the Joint IP should be clearly defined. Common delimiters are field of use and geography—the natural dividing lines are highly fact dependent. For example, a machine learning service provider might have the right to exploit improved algorithms outside the counterparty’s industry. A customer might also negotiate for a “head start” in using the improved algorithms in lieu of a full field prohibition on the service provider.
- Related to exploitation rights, the parties should consider whether the parties can license Joint IP to third parties (with or without consent of the party) and whether there is any obligation to share revenues with the other party.
- An often overlooked issue is the parties’ rights after termination of the collaboration. It is important to consider and plan for whether either party can continue to access the contributed IP for purposes of using the Joint IP or developing derivative IP.
1Vesta Corp. v. Amdocs Management, Inc., No.3:14-cv-1142HZ (D. Or. Jan. 13, 2015).
2Polyzen, Inc. v. Radiadyne, LLC, No. 5:11-CV-662-D (E.D.N.C. Sep. 23, 2016); see also Lucent Techs., Inc. v. Gateway, Inc., 543 F.3d 710 (Fed. Cir. Sep. 25, 2008) (finding Lucent lacked standing to sue because the patent at issue was jointly owned with Fraunhofer); Akzo Nobel Coatings v. Dow Chem. Co., C.A. No. 8666-VCP (Del. Ch. June 5, 2015) (denying motion to dismiss breach of contract suit after reviewing definitions of IP and IP ownership in the JV agreement).