Developing an IP Strategy for AI in a Rapidly Evolving Legal Landscape
Globally, many industries are experiencing intense competition as market leaders and would-be disruptors race to deploy artificial intelligence (AI) in products, services, and business processes. Driven by substantial investment and rapid innovation, AI tools are already being used to automate high-level tasks, enhance productivity, and guide decision-making across industry ranging from healthcare to transportation to construction, among others.
These investments in software and data science naturally raise questions on how best to protect AI assets. Obtaining intellectual property (IP) for AI assets can help protect market position or set partnership terms. IP can help in attracting interest of investors, substantiating freedom-to-operate assertions during due diligence, defining the scope of strategic licensing deals, deterring competitors from establishing/enlarging their market footprints, or mounting a response to a competitor’s IP infringement claims, and more. If a company is building or deploying AI, it should evaluate how IP will aid those goals.
While protecting other technologies may be relatively straightforward, protecting AI work can in some cases be complex. Given the technical and legal complexities, defensible and valuable IP protection rests on experienced counsel tailoring the IP strategy through a deep understanding of the product and business. For companies that leverage IP to advance business interests—rather than getting patents as trophies—a detailed discussion with counsel of the ideas, the competitive market, the business goals and product design is key to obtaining defensible and valued protection for AI.
Defensibility and Value of AI IP
For companies with well-established and robust processes for identifying and securing IP rights in valuable innovations, an adaptation or fine-tuning of those processes may be necessary to reliably obtain strong IP rights in AI assets. In the U.S., the legal landscape relating to obtaining and enforcing IP rights in AI innovations is already complex and continues to evolve rapidly. Nuanced and AI-specific considerations are thus arising.
Despite recent uncertainty about the patentability of some types of software, in both Europe and the U.S., leading judicial decisions and administrative guidance have affirmed that AI software and workflows are patent-eligible. In many industries, trade secrets are also an increasingly effective option for protecting AI assets—though disclosure requirements in heavily regulated industries may foreclose the possibility of relying on trade secrets to protect some AI assets. Nowhere is the phrase “data is the new currency” truer than with high-quality, anonymized data sets that companies gather and leverage in product development. Particularly when a product uses conventional hardware or runs on someone else’s hardware, obtaining IP for the AI, software, and data can be crucial for a company.
AI IP is also valuable. In October 2023, after prevailing on validity and infringement of its AI-related patents at trial, a smaller healthtech company obtained an order from the U.S. International Trade Commission blocking import of a major computer company’s smart watches. That larger company ultimately removed the health monitoring technology covered by the order from its watches, limiting competition for the smaller company. Likewise, one of the largest-ever trade secret cases centered on AI technology for use in autonomous vehicles, in which a company agreed to pay hundreds of millions of dollars following allegations a new employee brought trade secret material from a former employer. With a large multitude of AI data cases pending, over whether data used for training AI models was properly licensed, many companies are seeking to license their data to model developers, and model developers are eager for licensed data sets.
Avoiding Pitfalls and Securing High-Value IP
Against this backdrop of high value and concrete protection, companies can nonetheless face challenges. Some AI systems risk being labeled as unpatentable automations of pre-existing manual processes. Given that many of the fundamental AI techniques leveraged in today’s cutting-edge products (including “deep” learning) are years or decades old, there may be related prior work. For back-end functionality, detecting infringement can be tricky. And in some cases, specific AI techniques are short-lived and easily replaced by improved or alternative versions.
Rather than discouraging companies from protecting their AI assets, these challenges should inform the development of an IP strategy that adds value and avoids waste. Experienced IP counsel with expertise in AI can analyze these factors together with product details and business objectives, guide in-depth discussions necessary to align patents with value or supplement patents with other IP, and build a tailored IP strategy that helps the company achieve its goals.
Often, IP that provides the best protection for AI assets may not focus on the AI itself. Companies’ first thoughts are often of a specific model they designed or trained, but the exact model may be relatively short-lived. The model will be retrained and, in some cases, the model’s structure will be improved. While your team’s chosen model performs well, alternatives may be available, and detecting whether a competitor is using your model or another may be difficult. For certain high-value models, trade secrets may be appropriate, and patent strategy may focus more on other aspects of the product workflow that interact with the AI.
AI outputs typically drive downstream processing, leading to diagnostic output (e.g., in healthcare and cybersecurity), actionable predictions (e.g., in marketing and retail), control signals for closed-loop systems (e.g., in manufacturing, transportation, logistics, home automation, and energy delivery), recommendations to users, etc. Such workflows are fertile ground for AI IP. Though models evolve rapidly, AI-driven workflows may persist across product iterations. In contrast to models, these workflows are often highly visible to customers, tend to be the focal points of marketing campaigns and to drive customers’ purchasing decisions, and may be detectable when copied by competitors—all factors that increase patent value.
Data rights and licensing for those AI outputs also present high value, particularly in context with corresponding inputs. The data preparation processes that produce the inputs to AI models, such as data curation and pre-processing of raw data into informative features, may also be fruitful sources of IP. These input-output contexts may provide important support for patent protection. It also may be possible to consider trade secret protection where features or processing are not visible. Data rights for access to the input data can be priceless, particularly for training data obtained from partners. Such non-patent IP may be particularly important where the AI is largely mimicking a pre-existing manual process.
Next Steps
Developing an effective IP strategy for AI assets should involve detailed discussions with IP counsel of how the company is using AI internally and in customer-facing products and services to advance business goals. Such collaboration is key to determining which types of IP will provide the most defensible and valuable protection for each AI asset.
Author Bios
Andrew (A.J.) Tibbetts is a shareholder in the Intellectual Property & Technology practice group in Greenberg Traurig’s Boston office. He leverages prior experience as a software engineer to provide practical IP strategy counseling on matters related to computer- and electronics-implemented technology across a range of industries, including in healthtech, life sciences AI, computational biology, medical records analysis/coding, medical devices, and more. He can be reached at Andrew.Tibbetts@gtlaw.com.
Samuel S. Stone is an associate in Greenberg Traurig’s Boston office. He is an intellectual property attorney, a member of the firm’s Venture Capital and Emerging Technology Practice, and Innovation & Artificial Intelligence Group. He can be reached at Samuel.Stone@gtlaw.com.
LINKS
Read “Developing an IP Strategy for AI in a Rapidly Evolving Legal Landscape,” authored by Andrew (A.J.) Tibbetts for Lawyers Weekly. (subscription)