I’ve written previously about the Pentagon’s courtship of Silicon Valley. As military technologies grow increasingly reliant on software, communications networks, and autonomation, the need for innovative, fast-paced, and disruptive investments become all the more important. Leadership in the Department of Defense (DoD) has taken this to mean partnering with companies outside the traditional defense contracting sphere. The goal: procure next generation technologies and drive down costs by fostering expanded competition.
There is a major snag that could prevent the DoD from reaching this goal, and it is one that has often been discussed in defense circles: intellectual property. The military and the American tech industry have two very different approaches, and needs, when it comes to intellectual property. For companies, ownership of an idea often is the only thing standing between a billion dollar valuation and bankruptcy. For the military, the ability to replicate or unilaterally use an idea could be the difference between life and death. These different existential threats are hard to reconcile and will make the DoD’s mission to Silicon Valley tricky.
When the military releases requests for bids, especially in the software world, it includes language about the type of licensing rights it wants. For the military, the most useful licensing rights are either unlimited rights or government purpose rights. At the risk of getting overly technical, the important parts of these licenses are below:
- Unlimited Rights: the Government can use, modify, reproduce, release or disclose technical data or computer software in whole or in part, in any manner, and for any purpose whatsoever, and to have or authorize others to do so.
- Government Purpose Rights: rights to use, modify, reproduce, release or disclose the technical data or computer software within the Government without restriction and outside the Government for a Government purpose.
“Government purpose” is itself vague as well, and includes leeway for the government to transfer or sell products to international organizations or foreign governments. Both of these preclude commercial purposes, though it is easy to see why technology companies have problems working with the government. Once a product is sold to the government under these licensing rights, there is a loss of exclusive control over that product. For companies that have to quickly show the ability to generate profit, this might be a dealbreaker. If there is already a trust deficit, as there has been between Silicon Valley and the government after the NSA scandal, the willingness to turn over these rights to the government is even smaller.
In the most recent broad agency announcement from the Combating Terrorism Technical Support Office, the agency requested bids for advanced data visualization, as well as software to recognize desired people or objects within videos and pictures. Of course, companies like Google have been pursuing image recognition for years. But Google has also rejected working with the U.S. military for years. The lack of cooperation makes sense through the lens of the mentioned announcement. The military requires at minimum “government rights for the entire application.” In fact, “the system should be designed and developed so continual service and maintenance from the developer is not needed or required,” which is a problematic approach to an industry that essentially relies on continual business. It perhaps makes sense that the DoD would not want to get sucked into paying for legacy services, except that it already pays for legacy software at large cost.
The Pentagon’s approach to intellectual property makes it difficult for companies that do not already rely on government funding to get involved. Established defense contractors are already bought into the system, especially in providing military-specific equipment. When the federal government provided the majority of research and development (R&D) funding nationally, there were strong incentives for new companies to provide contracts. However, the federal government has provided less than 50 percent of R&D since 1979—meaning it likely remains more profitable for technology companies to maintain exclusive control of ideas and raise funds from the private sector. It is unlikely that the government can reverse this trend. In 2011, the federal share of American R&D was down to 30 percent— meaning, it can no longer rely on the size of its wallet to attract new ideas and contractors.
This problem will only be exacerbated in the future as new trends in technology emerge. 3D printing, for example, may make it even harder for companies to partner with the federal government. If a product can be stored as a piece of software, and then produced by any company or group with an applicable printer, contractors might bid for designs and lose the production part of the business. Companies might not abide the risk and prefer to only do business with the private sector. This already happens, as when Google purchased the robot manufacturer Boston Dynamics and vastly reduced its military contracts.
The Pentagon might argue that it is trying to keep its technology from proliferating into the commercial sector, and from there into the hands of enemies. This might be a useful approach for systems like fighter jets, intercontinental ballistic missiles, and nuclear warheads. However, for systems like data analysis and computers, the proliferation argument is flimsy. Many private companies will pursue these capabilities anyway, as they are integral to their own business. Private industry has already shown that it will fund these capabilities at a level equal to or higher than the Pentagon. These capabilities will exist in the commercial market anyway, and the question instead should be whether the Pentagon will spend valuable money replicating these capabilities— and likely falling short— or save money by working with the wider markets.
With technology becoming so much more integrated into civilian life, the private sector will likely continue to fund new developments and projects at a much higher level than the government. The Pentagon wants in on the results of these investments, both for new capability and for reduced costs. As of yet it is unclear what the private sector’s response will be. The success of the relationship, however, will likely rely more on the Pentagon’s ability to reform its approach to intellectual property and contracting, and less on how much it is willing to pay.