NSI Experts Weigh In: 2018 Department of Defense Artificial Intelligence Strategy

Yesterday, the Department of Defense (DoD) released a summary of its 2018 Artificial Intelligence Strategy.  Our experts weighed in about the strengths and weaknesses of this strategy. 

February 13, 2019

Bryson Bort – NSI Fellow; Founder & CEO, SCYTHE

“The DoD strategy cements the role of the Joint AI Center (JAIC) as the driver for AI development and adoption across all services.  Echoing the Administration’s AI directive, the strategy is framed in terms of the Chinese and Russian threats increasing through this technology.  In execution, there is a diametric tension between iterating slowly through ethical implications as evidenced by Google employees refusing to support Project Maven (a DoD AI project of imagery analysis) last year and the exhortation for ‘transformative AI technologies’ developed ‘from experiments at the edge.’  DoD has historically struggled with agile development that would support these ‘experiments’ from a contracting, cultural, and technical perspective.  The immediate wins look like the use of AI on the mountains of DoD data in various areas such as logistics and maintenance.”

Jamil N. Jaffer – NSI Founder and Executive Director; former Chief Counsel and Senior Advisor for the Senate Foreign Relations Committee and former Associate Counsel to President George W. Bush

“The new Artificial Intelligence Strategy released by the Department of Defense provides some unique insight into DoD’s well thought out plan on AI that it’s already been executing.  The specific outlining of the role of the Joint AI Center (JAIC) and the description of the key pilot projects that are ongoing, as well as the commitment to partnerships with leading private companies are key elements of a strategy that is likely to bear fruit in the short-term.  While much will be in the implementation, the early steps DoD has already taken are moving them in the right direction.

Harold Moss – NSI Visiting Fellow; Senior Director Strategy & Business Development, Web Products, Akamai Technologies

The Department of Defense’s recent strategy is a thoughtful approach to embracing an emerging technology.  By introducing Artificial Intelligence as an augmentation technology and building upon the proven techniques and skills used by the DoD, they can rapidly introduce the technology without introducing disruptions to our current national security strategy.  Furthermore, by leveraging a number of diverse sectors to usher in the new technology, they reduce the integration challenges found in emerging technologies while enabling the latest information to drive implementation.

I am especially encouraged by the strategies identification of focal areas that not only introduce meaningful impact, but rely on cultivating supporting internal/external resources and infrastructure. By implementing an approach which evolves with the pace of technology, the strategy will be far more sustainable, and at the same time introduce ongoing benefits to national security.

That said, while the strategy has expertly focused on how to introduce AI into the working models of the agency, they will need to establish relationships with foreign and domestic partners to establish the corpus of knowledge essential to AI.  As a result, the DoD will need to clearly articulate and enforce privacy and security protocols, which will establish confidence in partners and the program.”

Adam Pearlman – NSI Visiting Fellow; former Associate Deputy General Counsel, U.S. Department of Defense 

“Beyond the substance of the DoD’s AI strategy, the timing of this document is not to be overlooked.  For all the ways in which the White House has been criticized for going off-the-cuff, the issuance of Executive Order 13859 and the release of this document all within a week of the State of the Union, shows not only the importance the President places on this issue, but that efforts are being properly coordinated throughout the government.  This shows a maturity in managing the bureaucracy, and vastly increases the likelihood of building an effective capability.

It is also notable that a relatively sizable portion of the strategy emphasizes private-sector partnerships, which will be critical to the success of building the government’s AI capabilities.  We are long past the era when DoD had both more money and more information than everyone else.  The five most valuable tech companies have a market capitalization of more than three times the defense budget and are sitting on close to $600 billion in cash.  They also have incredibly large and diverse data sets that will prove critical to a cutting-edge AI system.  Recruiting all the engineering talent in the world is meaningless without having useful data to train an AI system.

Developing this technology while protecting civil liberties is no easy task, but having articulated guiding principles will help the Department and the government navigate these issues in a responsible way.”

Disclaimer: The views and opinions expressed in this analysis are those of the authors and do not necessarily reflect the official policy or position of the National Security Institute or any agency of the U.S. government. Assumptions made within the analysis are not reflective of the position of the National Security Institute or any U.S. government entity.