TLDR: A research paper by Andrew Bowne explores why the US Department of Defense (DoD) struggles to attract commercial AI firms despite being an attractive customer. It finds that traditional DoD contracts are slow, costly, and rigid, clashing with the agile nature of AI development. The paper proposes ‘Optimal Buyer Theory,’ advocating for flexible, collaborative contracts like Other Transaction (OT) agreements, better education for DoD personnel on AI and alternative contracting, and legislative changes to broaden OT authority. The goal is to align DoD practices with commercial preferences to secure critical AI innovation for national security.
In an era where artificial intelligence (AI) is rapidly reshaping industries and national capabilities, a critical challenge faces the United States Department of Defense (DoD): how to effectively attract and partner with the innovative commercial AI firms that are driving this technological revolution. A recent doctoral thesis by Andrew Bowne delves deep into this very issue, proposing a new framework to help the DoD become a more appealing customer to these crucial technology providers.
AI is not just another technology; it’s a disruptive force that augments human thought and action across countless domains. Unlike previous military technologies often developed with federal funding, AI advancements are predominantly fueled by commercial industry for civilian applications. This creates a unique situation for the DoD, which historically has been the primary driver of defense technology. Now, it finds itself competing with a vibrant commercial market for top talent and cutting-edge solutions.
The research highlights that the DoD faces significant hurdles in engaging with commercial AI firms. Traditional defense contracting, governed by the Federal Acquisition Regulations (FAR), is often perceived as slow, bureaucratic, costly, and rigid. This system, designed with transparency and integrity in mind, can inadvertently create high barriers to entry for agile, commercially-focused AI companies. Issues like lengthy contract award timelines, complex accounting standards, and inflexible intellectual property (IP) clauses deter many innovative firms who find commercial markets easier and more profitable to navigate.
However, the study reveals a paradox: despite these challenges, commercial AI firms generally view the DoD as an attractive customer. This attraction stems from several factors, including the potential for significant revenue, the credibility a DoD contract lends in the commercial market, the opportunity to work on challenging and meaningful projects, and a sense of patriotism or desire to contribute to national security. The professionalism and mission-oriented nature of military personnel also stand out as positive attributes.
The research identifies two primary legal frameworks for DoD contracts: the traditional FAR-based contracts and the more flexible Other Transaction (OT) agreements. The study strongly suggests that OT agreements are far better suited for acquiring AI capabilities. OTs offer streamlined processes, encourage open communication and collaboration with end-users, allow for flexible negotiation of terms (especially IP rights), and support agile, iterative development—all crucial aspects of modern AI development. In contrast, FAR contracts, with their fixed requirements and extensive mandatory clauses, are often incompatible with the dynamic and experimental nature of AI innovation.
A key finding is that commercial AI firms prioritize collaboration, communication, and flexibility. They prefer to be involved in defining the problem and iterating on solutions, rather than simply following pre-set specifications. Intellectual property protection is also paramount, with many firms viewing their IP as their most valuable asset. The ability to negotiate IP terms that align with commercial practices, rather than being bound by rigid government-centric clauses, is a major draw.
To address these issues, Andrew Bowne introduces the “Optimal Buyer Theory.” This theory posits that the DoD can maximize its attractiveness to commercial AI firms by making deliberate choices in its contracting process. An optimal buyer fosters collaborative relationships, characterized by transparent communication, flexibility for experimentation and iteration, negotiated terms that benefit all parties, and a clear purpose-oriented approach. This means moving away from purely transactional contracts towards partnerships built on trust and mutual goals.
The thesis offers several recommendations for the DoD. Firstly, it should implement best practices in contracting by leveraging existing legal tools like OT authority more effectively. This involves prioritizing flexibility, negotiation, and collaboration. Secondly, the DoD needs to invest in education and training for its workforce, not only in OT agreements but also in understanding commercial contracting practices and the unique complexities of AI technology. This will empower contracting professionals to navigate complex negotiations and tailor contracts to attract innovative firms. Thirdly, Congress should consider broadening the DoD’s authority to use OTs, potentially mirroring NASA’s more expansive Space Act Agreements, to further reduce statutory limitations. Finally, the research suggests further study into how internal factors, such as budgeting systems, impact the DoD’s ability to attract commercial AI firms, and advocates for continued collaboration with industry, academia, and international partners to develop a common ethical framework for AI. You can read the full research paper here: Attracting Commercial Artificial Intelligence Firms to Support National Security Through Collaborative Contracts.
Also Read:
- Bridging Disciplines: The Essential Role of Interdisciplinary Collaboration in AI Mental Health Chatbots
- The Looming Divide: How Autonomous AI Agents Could Create New Societal Disparities
By embracing these changes, the DoD can transform its contracting approach, ensuring it gains timely access to the cutting-edge AI innovation necessary to maintain national security and compete effectively in the AI era.
