The Department of Defense (DoD) is facing an unprecedented challenge: harnessing the power of artificial intelligence (AI) to maintain a competitive edge in an increasingly complex world. At the forefront of this endeavor is Craig Martell, the outgoing chief digital and artificial intelligence officer (CDAO) of the DoD, whose insights offer a compelling vision for the future of AI within the department.
Martell’s departure marks a critical juncture, where the DoD must confront the reality that the lion’s share of AI innovation now originates not within its own walls, but in the dynamic ecosystem of private technology companies. Martell works to underscore the need for a fundamental shift in strategy, one that recognizes the pivotal role of private industry in shaping the DoD’s AI capabilities.
Unlocking the Power of Data
Central to Craig Martell’s vision of the Department of Defense’s AI future is the recognition of data as the foundation of effective AI applications. Martell, in his role as the outgoing chief digital and AI officer of the DoD, emphasizes the critical importance of high-quality data in driving AI innovation.
In Martell’s view, approximately 80% of the demand for AI can be met with high-quality data and effective dashboarding systems. This assertion highlights the notion that without access to clean, accessible, and usable data, the potential of AI remains largely untapped. Martell also stresses that the true value of data lies not merely in its volume, but in its utility, and to be valuable data must be actionable and meet the specific needs of end-users, particularly combatant commanders.
However, Martell acknowledges that a significant hurdle: much of the DoD’s data remains inaccessible or improperly formatted, rendering it effectively unusable for AI applications. To address this challenge, Martell advocates for a paradigm shift in how the DoD approaches data utilization. Rather than relying solely on internal efforts to curate and manage data, Martell proposes creating an environment conducive to private contractors and services developing AI applications using the vast troves of untapped data stored within the department.
Martell’s approach emphasizes the removal of limitations that impede data usability, particularly through initiatives such as data labeling. He argues that labeling data, though often overlooked, is a critical component of data quality and represents a form of intellectual property that holds immense value. Drawing from his extensive experiences in the private sector, Martell highlights the arduous nature of data labeling, noting that it often consumes the majority of the researcher’s time, a sentiment echoed by many in the field.
To overcome these challenges, Martell envisions the evolution of data labeling into a service-oriented model, where data labeling becomes more accessible and streamlined. He anticipates the forthcoming release of “data labeling as a service” as a significant step toward achieving standardized, widely available data for AI applications.
Transitioning to Modern Practices
Martell, drawing from his extensive experience and insights gained during his tenure, advocates for a fundamental shift in the DoD’s approach to contracting and policy frameworks. He highlights the inherent limitations of outdated practices dating back to the 1960s when defense dollars were the primary driver of American innovation and the private sector’s role was relatively limited.
To address these shortcomings, Martell calls for agile, responsive, and fast-moving contracting and policy mechanisms. He emphasizes the importance of modernizing these frameworks to align with the rapid pace of technological change and enable the DoD to leverage the latest advancements in AI from the private sector.
Kathleen Hicks, deputy secretary of defense, echoes Martell’s sentiments, focusing on the need to update contracting standards to current industry norms. Hicks emphasizes the necessity of building contracting and policy frameworks that are adaptable to technological advancements, ensuring that the DoD remains at the forefront of AI innovation.
Martell cites examples of successful implementations of new contracting systems under his office’s leadership, highlighting the ability to onboard new vendors in a matter of weeks rather than months or years. He emphasizes the importance of establishing clear guidelines and rules of engagement for vendors, enabling them to navigate the procurement process with confidence and agility.
Martell also stresses the importance of fostering an environment where vendors can innovate freely while adhering to established standards and regulations. He envisions a future where AI scaffolding within the DoD is designed to accommodate diverse vendors and technologies, providing flexibility without sacrificing security or efficiency.
Encouraging Innovation while Ensuring Transparency
The DoD has recognized the pivotal role of private industry in driving AI innovation, clearly seeing that it cannot compete with major corporations in terms of resources or talent. However, there is an imperative for transparency and openness in the development and deployment of AI technologies for national defense.
A key aspect of Martell’s vision is the need for vendors to adopt new revenue models that prioritize transparency over proprietary control. He asserts that vendors offering novel AI tools to the DoD should allow the department to have full visibility into the algorithms employed, ensuring accountability and mitigating potential risks.
Advocacy for transparency extends beyond the technical aspects of AI development to encompass broader principles of ethical and responsible use. He emphasizes the importance of open dialogue and collaboration between government, industry, and academia to address ethical concerns and ensure that AI technologies align with national security objectives and societal values.
There is an apparent need for clear guidelines and regulations to govern the use of AI within the DoD, balancing the imperative for innovation with the necessity of accountability. He envisions a future where vendors adhere to established standards of transparency and accountability, enabling the DoD to leverage the transformative potential of AI while minimizing risks and ensuring compliance with regulatory frameworks.
Balancing Standards and Flexibility
While the pursuit of standardization and regulation is essential for success with AI, flexibility will also play a central role for government bodies such as the DoD. Martell emphasizes the importance of empowering stakeholders to tailor AI solutions to their specific needs, allowing for innovation and customization while still adhering to overarching standards and guidelines.
Martell proposes a future where the DoD maintains a balance between enforcing standards for joint operations and allowing flexibility for service-specific initiatives. There will be a clear delineation between core standards that ensure interoperability and flexibility in areas where customization is warranted, enabling each branch of the military to optimize AI capabilities to suit its unique mission requirements. To achieve this, there is a need for continuous evaluation and refinement of standards to establish relevance and effectiveness in a rapidly evolving operational environment. He advocates for an iterative approach to standards development, incorporating feedback from stakeholders and adapting to emerging technologies and threats.
Building a Skilled Workforce
Central to the successful integration of artificial intelligence within the DoD is the cultivation of a skilled and adaptive workforce. Investing in human capital is a critical component of AI readiness and capability development. There is scarcity of AI expertise within the DoD compared to the private sector. However, Martell remains optimistic about the department’s ability to build a capable workforce skilled in leveraging AI to its fullest potential. There is a need to adopt a proactive approach to talent development, focusing on training and equipping practitioners with the requisite skills and tools necessary to drive innovation.
Members of the government have advocated for a strategy that revolves around training AI practitioners within their own programs to instill the knowledge and resources they need to succeed in their roles. Providing targeted training programs tailored to specific job roles and skill levels ensures that practitioners receive relevant and practical instruction that aligns with the evolving demands of AI integration.
This approach to building a skilled workforce within the DoD is rooted in the recognition that talent is the lifeblood of AI innovation. It creates a culture of continuous learning and adaptation, where practitioners are encouraged to live at the forefront of AI development and deployment.
