One of the main recommendations from the National Security Commission on Artificial Intelligence’s final report released in 2021 was that the U.S. Department of Defense must be AI-ready by 2025. This guidance was based on concerns that China was moving rapidly to overcome U.S. technological advantages and the impact this maneuvering would have on the DoD and the defense of our nation.

Since the NSC report was released, the DoD and military services have sought to make this recommendation a reality. They have undertook a number of highly publicized efforts directed at acquiring AI technology and creating relevant data models for machine learning to bring these capabilities to the warfighter community.

However, the DoD and the military services are still moving slowly to becoming AI-ready and I wonder how serious their intentions are.

DoD AI Pilot (Lack of) Success

I recently engaged with leaders of a well-know company that provides the Federal government with AI solutions. The leaders shared that their company has tracked 1,300 AI pilots in the government (70 percent percent in the DoD alone), but a mere 10 percent have successfully moved from trial to production. And of the DoD’s estimated 910-plus pilots, only 91 have progressed to deployment.

While many reasons can be cited for these poor success metrics, studies by academia, research institutions and private-sector companies have clearly outlined a top culprit: a lack of AI and data literacy.

Literacy Is the Key for Successful AI and Data Projects

Ongoing training programs are absolutely necessary for employees to be able to do their jobs efficiently and effectively—particularly in the evolving world of work impacted by technological innovation. The same most definitely applies to AI and data science. A workforce with the right knowledge and skills is critical for AI and data project success. This assertation is supported by a multitude of research on the topic from academia, research institutions and corporations. As follows is what a few studies have to say about the connection between AI and data literacy and successful projects in these areas:

1. The Impact of Data Literacy on Project Success. A study by Ponemon Institute (2019) found that organizations with a higher level of data literacy are more likely to achieve success in their data analytics projects. Data literacy leads to more effective communications and better, or “smarter,” interpretation of data insights, which ushers improved decision making and project outcomes.
2. The Importance of AI Literacy in AI Projects. This study by MIT Sloan Management Review (2020) discovered that organizations that invest in AI literacy are more likely to succeed in their AI projects. A lack of literacy among team members can cause the misinterpretation of AI outcomes—negatively affecting project success.
3. The Role of Data Literacy in AI Ethics. A study by Deloitte (2020) showed that data literacy is essential for ensuring ethical, or responsible, AI. The study revealed that organizations with higher levels of data literacy are better equipped to understand the ethical implications of their AI projects, which can improve project outcomes and mitigate risks.
4. The Impact of Data Literacy on Employee Performance. This study by Accenture (2018) shared that data literacy improves employee performance and project outcomes. Organizations higher levels of data literacy have more engaged workforces that are better equipped to understand data insights and make data-driven decisions.
5. The Importance of AI Literacy in Executive Decision-Making. A study by Harvard Business Review (2019) found that AI literacy among executives is crucial for successful AI projects. Executives possessing higher levels of literacy are better equipped to understand the potential of these advanced technology solutions and make more informed decisions about AI projects.

What Is the DoD Doing About AI and Data Literacy?

In short, I believe that the DoD is not doing nearly enough about educating its workforce in AI and data science…with the caveat that I also understand the magnitude of this undertaking.

The former DoD Joint AI Center and the current Chief Digital and AI Office were provided responsibility to begin the process of reskilling and upskilling the Department’s workforce (both civilian and military). Their efforts to date are commendable. They have piloted a number of AI training courses across the gamut of DoD military and civilian senior leaders to end users.

The military services also have initiated AI training pilots to upskill and reskill their workforces. For example, the U.S. Army recently announced a “train-the-trainer” program to provide foundational AI and data literacy knowledge and skills to volunteers who then can provide instruction to their units. The U.S. Air Force also created Digital University, where Air Force, U.S. Space Force and DoD personnel can access relevant online training opportunities from third-party providers. Another example in the military services is offering coding bootcamps to their personnel. However, I must note that the vast majority of these initiatives and others are focused on instructor-led training—an expensive delivery method that is not scalable.

Priority for AI and Data Literacy Training

The challenge to provide AI and data literacy training to the DoD workforce is staggering. For the DoD to achieve its goal of becoming AI-ready, all 3.2 million military and civilian personnel must receive some level of upskilling or reskilling. To date, the DoD and military services have emphasized training for senior leaders and acquisition and technical workforce roles (e.g., AI/machine learning engineers, software engineers, data scientists)—or approximately 500,000 of the total workforce. This means that approximately 2.7 million employees—end users of AI solutions working in operations, intelligence, logistics and maintenance, health, human resources and more—still need some degree of basic AI and data literacy education.

These 2.7 million people—the high majority of the DoD workforce—should be the priority for AI and data literacy training. The Department’s race to AI-readiness depends on it.

Dr. J. Keith Dunbar
Founder and Chief Executive Officer
FedLearn