The Military’s Insistence on Owning Commercial Intellectual Property is Limiting Innovation
In early December, the Department of Defense Intellectual Property Cadre hosted its inaugural “DoD IP Forum,” where government and industry leaders convened to debate the growing role of commercial software in national security and how the U.S. government can improve its intellectual property posture towards industry to achieve better outcomes for the department.
At first glance, the issue of commercial intellectual property rights may seem unrelated to how the U.S. government seeks to meet the needs of America’s warfighters, but Department of Defense leaders are beginning to recognize that intellectual property policy has become a serious obstacle to defense innovation. In the age of advanced software and AI, the U.S. government is coming to rely more on the procurement of dual-use commercial technology to give the armed forces a strategic, operational, and tactical edge. In fact, in the first-ever National Defense Industrial Strategy released last week, the Department of Defense explicitly highlighted the importance of intellectual property and data rights policies as a pathway to flexible acquisition and sustainment.
Unfortunately, today, the U.S. government and the commercial software industry are frequently at odds in their vision for how much government intellectual property ownership is required for critical national security programs to succeed. There is a requirement for the secretary of defense to “define the legitimate interest of the U.S. government” when asking companies to give up their intellectual property. While this requirement is meant to protect commercial intellectual property, in practice many commercial vendors are subject to solicitation requirements for either “unlimited rights” or “government purpose rights.” Unlimited rights enable the government to use, modify, reproduce, release, or disclose technical data or computer software in whole or in part, in any manner, and for any purpose whatsoever, including to a company’s competitors. Government purpose rights are slightly more restrictive, but still allow the government to use the software within government without restriction and outside government if there is a government purpose.
The result is that software providers eager to support the Department of Defense risk losing the very assets that sustain their business: company intellectual property. This not only dissuades commercial technology providers from seeking government contracts, it places a particularly onerous burden on younger non-traditional tech firms that are hoping to break into the defense industrial ecosystem — both of which limit the public-private partnerships necessary to ensure that the United States stays ahead of technologically advanced competitors.
As senior leaders at Palantir Technologies and Tangram Flex, software companies that do a great deal of business with the Defense Department, we have a direct interest in contracting outcomes. But we also bring a unique perspective that is formed by a cumulative 40 years of expertise working through the barriers of commercial software adoption by the U.S. government. Based on our experience, we believe that for the Department of Defense to vastly improve the speed and scale with which it acquires and fields emerging technology, it should rethink its current approach to intellectual property ownership.
Specifically, the government should stop conflating intellectual property rights with vendor lock and interoperability. One does not solve for the other, and thus commercial companies should not have to forfeit their core intellectual property — namely patents, copyrights, trademarks, and trade secrets — in the name of interoperability. This approach can serve as a win-win for defense and industry by effectively addressing the root cause of interoperability and vendor lock issues, incentivizing commercial companies to provide their most innovative solutions to the government and boosting the expansion and health of America’s budding defense tech start-up ecosystem.
The Current Landscape
What’s at stake in the data rights debate? What are potential points of tension between government and industry? And what opportunities exist to ensure mutual benefit that supports America’s warfighters?
What Does Government Want Out of Data Rights?
Data rights is a broad category that refers to both a company’s intellectual property and ownership of government data. While ownership of government data is self-explanatory, for software firms, intellectual property refers to the proprietary source code needed to recreate the software. Enterprise software-as-a-service companies are not advocating to own customer data, believing that the U.S. government should always exclusively own its data. Instead, disagreements between government and industry tend to revolve around rights to commercially financed and owned intellectual property.
It is critical that the Department of Defense be able to own and access its data for many reasons, including preparedness and the execution of war, the sustainment and improvement of existing capability, senior leader decision support, and more. As the 2020 Department of Defense Data Strategy explains, “DoD data is a high interest commodity,” and it “must be made available for use by all authorized individuals and non-person entities through appropriate mechanisms.”
While the department’s requirement to exclusively own its data is both reasonable and necessary, the claim to commercial intellectual property is more complex and controversial. Within the department, there is a general view that owning, not renting, a new capability or system — through requiring unlimited rights and government purpose rights requirements — will help avoid vendor lock with specific software providers and improve the interoperability of software-based systems. This term refers both to the government being locked into using a specific vendor (and therefore being susceptible to price extortion), as well as being locked in to a given vendor’s architecture that is not interoperable with the government’s current or future investments.
For example, the Department of Defense’s 2022 State of Competition in the Industrial Base concluded, “If a component or subsystem becomes vendor-locked, the process for acquiring, supporting, and upgrading the overarching system or subsystem can be subject to the pressure of that vendor lock, because the overall support for that system depends on the exclusive IP rights governing a critical subsystem or component.” While the department’s fear of vendor-lock tends to focus on software providers (versus hardware providers), the concern is universal. After all, the production of a plane, tank, or ship can include substantial vendor-lock as the government does not have the opportunity to switch providers until there is a new class of that platform.
Simply put, vendor-lock is the opposite of interoperability: It limits competition for government contracts, increasing costs; it undermines the government’s ability to engage in flexible modification and sustainment when software updates are needed, or during times of rapid change in requirements; and it ultimately prevents the government from acquiring the best new technologies.
What Does Industry Want Out of Data Rights?
While commercial software companies do not seek ownership over government data, the request by government buyers for unlimited rights and government purpose rights requirements on core commercial intellectual property is a difficult proposition. However, this preference is not related to commercial providers seeking to create vendor lock-in or undermine interoperability. In fact, we advocate that industry consent to unlimited rights or government purpose rights on non-core intellectual property, including all derivative logic, models, analysis, interface control documents, application programming interfaces to other government systems, and computer software configuration items.
Commercial software companies object to the sweeping terms stipulated in unlimited rights and government purpose rights because protection of core intellectual property — typically built with independent research and development dollars or outside funding like venture capital — is central to their ability to remain a viable company. If companies had to give up their intellectual property every time they sold to a customer, whether government or commercial, they would struggle to remain in business as it would be too easy for competitors to copy their product for free. American tech companies don’t want their intellectual property stolen by China, and they also don’t want to turn it over to the U.S. government as the price of entry to provide national security capabilities.
Why the Defense Department’s Approach Isn’t Working
While the battle for intellectual property ownership is perceived as existential by both the Department of Defense and commercial providers, a closer look at the origins of each actor’s preferences reveals greater compatibility. Industry wants to ensure a return on its investment, which doesn’t require promoting vendor lock or undermining interoperability. The military wants to limit vendor lock and promote interoperability, regardless of whether commercial firms earn a return on investment. The key question then is: Does department ownership of commercial intellectual property actually support its stated interests, justifying the harm it causes to commercial actors? The answer, in short, is no.
First, intellectual property ownership is not actually necessary to combat vendor lock and interoperability. For example, even if the department does not own a vendor’s baseline intellectual property, a legal precedent already exists so that no vendor can preclude the government from cloning the application programming interfaces of the function it wants and then re-competing the capability to the market.
Additionally, the issue driving vendor lock and high switching costs is proprietary interfaces, not proprietary code or intellectual property. As a former special advisor to the secretary of the Air Force, Tim Grayson, said about software solutions from traditional defense contractors: “When we have worked interoperability at the data level what we have found is the biggest problem is legacy programs and their vendors don’t know their own systems. We might have to have them spend three to six months effectively reverse engineering their systems and documenting what their interfaces are.” Ironically, the complexity and bespoke nature of many custom government off-the-shelf solutions drives outcomes that the government is trying to avoid: increased technical debt and vendor lock to the developer.
Second, regardless of who owns the intellectual property, the military can mandate application programming interfaces and interoperability with relevant government reference architectures. The department can also write into contracts mandatory data export or “burn down” requirements for key deliverables, which lowers the switching costs of changing vendors. This should be sufficient to meet most interoperability standards, and the department can always insist on a re-compete for any capability it feels does not pass muster.
Third, the Department of Defense has already authorized new approaches to maximize interoperability and avoid vendor lock. For example, in 2017, U.S. law mandated that all major defense acquisition programs must be designed and developed using a modular open systems approach, and a 2019 tri-service memo reinforced this approach. The modular open systems approach is a technical and business strategy designed in part to lower the switching costs of moving from one vendor to another by mandating open standards and interoperable systems. Unfortunately, when the Department of Defense encourages unlimited rights or government purpose rights in the name of interoperability or a modular open systems approach, they box out commercial companies, who, ironically, are the entities most likely to be compliant with this approach. By definition, dual-use commercial capabilities are trying to capture a large segment of the total addressable market, and thus commercial providers build their products to be as open and modular as possible to satisfy the widest possible set of requirements.
Fourth, we have heard officials express fears that during wartime conditions — where urgent changes and modifications must be made — the government will be unable to make system changes because it doesn’t own the source code and thus becomes reliant upon commercial actors. However, the ease with which a program can use and modify a technology is almost entirely divorced from “owning the source code.” To use a non-defense example, Windows users cannot modify their operating system source code, but there are limitless configurations and customizations they can build on top of this commercial baseline. Additionally, even if the U.S. government owned the source code, it likely would need a more upskilled workforce to modify, test, and deploy changes to the source code of a company’s commercial baseline under acute stress.
Finally, the requirement of unlimited data rights or government purpose rights for commercial software is resulting in the U.S. military missing out on innovation. When faced with such requirements, commercial companies are either forced to opt out completely, or they create a special-purpose version of their software with the desired rights — this version, or branch, of software is frozen in time from that point on. Companies will then release new features and innovations only in newer versions of their commercial software, a necessity for them to remain competitive in the marketplace and protect their product development investment. The Department of Defense misses out on the key benefit of buying commercial software — frequent new releases with additional features driven by feedback from a large customer base.
Over the past several years, the U.S. government has begun to wield new authorities, like the modular open systems approach, to avoid vendor lock and improve interoperability (perhaps an implicit acknowledgment that aggressive data rights are not yielding desired outcomes). Yet even with these new tools, the military persists on unnecessary claims to unlimited rights. The result is the worst of all worlds — interoperability issues persist, and commercial firms become disincentivized from pursuing the defense tech market altogether.
Recommendations
Given that the military can maximize interoperability without claiming maximum intellectual property rights, how can the Department of Defense and industry work together to ensure American warfighters have access to the best capabilities American industry has to offer?
First, and foremost, we recommend that procurements allow commercial software vendors to retain intellectual property rights to their commercial baseline. Second, to procure solutions that are configurable, flexible, and responsive — especially in wartime — the military should shift its evaluation criteria for software procurement from a heavy focus on intellectual property ownership to the outcomes of interoperability and responsiveness. The F-35 program is a clear case where focusing on data rights — as opposed to interoperability directly — lead to worse outcomes for the military. In this case, the government is on the hook to pay $500 million for Lockheed Martin to export data from the F-35, with the F-35 program executive admitting, “It’s not a matter of data rights, it’s matter of data delivery.” Despite the U.S. government having its desired data rights, it still has a dependency on Lockheed Martin for the data delivery needed to sustain the aircraft.
Third, to ensure vendors and capabilities will be sufficiently responsive in crisis situations, programs should include parameters and inputs that the program must be able to tune autonomously to be responsive to emergent needs, as well as service-level agreement requirements for modifications and new features. Programs can then hold a competition and see if commercial software companies, or government off-the-shelf providers, can ship features faster during operational exercises.
Commercial providers also have a responsibility to create a conducive environment for policy change. First, as there is still a strong cultural conflation between commercial intellectual property and government data ownership, industry should proactively demonstrate the ability for government customers to access and export their data outside of a company’s product, as well as be willing to be held to contractual obligations to the same effect. Second, companies should spend more time building trust with their Department of Defense customers by explicitly demonstrating how commercial companies build their product to be as open and interoperable as possible, for example with application programming interfaces, micro-services, and the use of open source components.
The benefits of acting on these recommendations will be clear. In addition to improving interoperability for military programs, allowing commercial software providers to own their intellectual property will help keep America’s industrial base strong, providing the best capabilities available to national security community. As a recent George Mason University report argued, “if DoD wants the best commercial firms solving military problems, it will have to respect their primary asset — intellectual property.” Furthermore, as a 2022 National Defense Industrial Association report asserts, “Intellectual Property (IP) rights are essential to the health of the [defense industrial base]. The perception of risks to IP rights shapes investor’s willingness to invest in research and development and commercialization activities.” This matters because the private sector invests in research and development at 3.5 times the rate of the U.S. government, and commercial software has a large customer base that helps drive innovation and new functionality.
Conclusion
The U.S. government has won a Pyrrhic victory with its current posture on data rights. It gets the intellectual property it wants but not the interoperability and innovation it needs. Commercial companies are put in an untenable position and boxed out from providing their best capabilities to the nation’s hardest problems. If the U.S. government mandates interoperability requirements and protects the baseline intellectual property of commercial software, there is a win-win for all parties involved.
Importantly, this creates opportunities for the next generation of defense tech innovators who are keen to break into the defense domain. Unnecessarily onerous intellectual property requirements have been stifling an entire generation of budding defense tech firms eager to support the Department of Defense. By allowing these young firms to keep their intellectual property, it improves their ability to stay competitive, earn return, and most importantly contribute their innovative solutions to America’s warfighters.
Shyam Sankar is the Chief Technology Officer of Palantir Technologies.
Caitlin Dohrman is the Chief Executive Officer of Tangram Flex.
Madeline Zimmerman is a Department of Defense Deployment Strategist at Palantir Technologies.
Image: U.S. Air Force photo by Senior Airman Quion Lowe
Comments are closed.