Artificial intelligence (AI) and machine learning are transforming how transportation solutions are built and deployed. To make advancements toward fully autonomous vehicles, strategic collaborations are needed to bring together appropriate technological expertise and data resources. However, such alliances require careful consideration of how ownership and control of intellectual property (IP)— including data—should be allocated, and how IP and data risk should be shared between the collaborators.
Autonomous vehicles (AV) have the potential to transform transportation in the near future, and many car manufacturers are adopting AI and machine learning to embrace these possibilities. Collaborations between Toyota and Uber and chipmaker Nvidia and Volkswagen are part of a growing trend of car manufacturers and technology companies working together on next-generation AV technologies. Such collaborations merge synergistic technologies and IP, and promise to accelerate the rate of advancement in the AV industry.
Furthermore, AI and machine learning are often deployed collaboratively in transportation because they require a variety of data inputs from different parties. For example, a wide range of private and public data sources are needed to develop and train the AI algorithms and machine learning models that function as the brains of AVs.
Strategic collaborative efforts can unlock significant value for the parties involved. However, engaging in collaborations requires careful consideration of several important legal issues.
One key issue is the allocation of IP rights between the collaborators. These rights may encompass, for example, those in inventions, such as driving algorithms or ethical AI frameworks. They may also include rights in data, such as training data, machine learning model parameters and other types of derived data, all of which may have tremendous economic value.
Allocation of intellectual property rights between collaborators requires analysis of what IP each collaborator brings to the table—often referred to as background IP—and the business objectives of each party, which are often not simply monetary and include exploiting collaboratively developed technologies and IP. This analysis feeds into the negotiation and preparation of bespoke contractual agreements that govern how IP is pooled during the collaboration, and to what extent each collaborator can enjoy the fruits of the partnership once it ends.
Failure to put in place proper contracts may result in a collaborator losing access to new technologies created during the collaboration, or even losing control over its own background IP. Lack of proper contractual protections may also result in undesirable fragmentation of IP rights between the parties, such that neither participant can effectively exploit collaboratively developed technologies. These potential pitfalls underscore the need for carefully crafted agreements and associates’ rights allocation frameworks that ensure collaborators can achieve their respective strategic aims.
Another key issue is the allocation of risk between the parties. For example, collaborators will need to decide which of them bears, and to what extent, the risk and the potentially large cost of defending against assertions of third-party IP. There is also significant regulatory risk with respect to data privacy, and of course, there can be substantial human risks if autonomous vehicle systems fail.
Another important set of risks is related to any gaps in end-to-end data compliance: How and under what associated legal terms has data has been sourced, and has the collection and processing of this data complied with applicable legal and regulatory standards? Data collection and processing tends to cross borders, and associated laws are evolving rapidly, with new jurisdictions announcing not only new data privacy laws, but also planned laws regulating specific technologies relevant to AV.
Properly allocating risks with a contractual agreement often involves placing particular types of risk on the shoulders of the party best positioned to mitigate that risk or weather potential economic fallout. Improper allocation of risk may leave collaborators vulnerable to risk outside their control and expose them to significant economic losses.
Various types of agreements governing collaborations can be used to address the legal issues highlighted above, and part of preparing a bespoke contractual agreement or framework involves selecting the right form of agreement based on the nature of the co-operation contemplated by the parties. For example, in more arm’s-length collaborations, IP licensing or sharing agreements may be appropriate. However, when parties intend to combine their expertise to develop new technologies, co-development agreements may be the right choice. And when parties intend to bid together for a contract, a teaming agreement may be the way to go. Both co-development and teaming agreements allow collaborators to pool their resources and work together toward a defined purpose. When the parties prefer to establish closer ties, for example, in order to balance sharing of risks and rewards, other arrangements such as partnership or joint venture agreements may be appropriate.
Given the large number of technologies and collaborators involved in AV, and the value of agile contracting in driving innovation forward, it is worth investing in contracts that not only reduce friction points in getting deals done, but anticipate and help manage failure points if collaborations or technologies don’t meet planned objectives. Creative methods for steering parties toward balanced approaches based on “market terms,” clear and simple IP and data governance approaches, and mediation of disputes are examples of key considerations in developing a workable contract framework for AV collaborations. Tools and guidance to aid counterparties in aligning commercial and IP interests, and document automation are also essential.
The key take-away is the need for bespoke contractual arrangements that govern collaborations. Parties should use the right form of agreement and negotiate terms to allocate the risk and rewards of the collaboration so they can pursue their strategic aims while being shielded from undue risk.
The authors would like to thank Norton Rose Fulbright Canada associate Saba Samanianpour for her assistance in preparing this article.