Pentagon unveils long-awaited plan for implementing ‘responsible AI’

Deputy Secretary of Defense Kathleen Hicks signed the Responsible Artificial Intelligence Strategy and Implementation Pathway (RAI S&I pathway) on Tuesday marking a highly anticipated next step in the Defense Department’s carrying out of its AI Ethical Principles adopted more than two years ago.

The 47-page document directs the sprawling Pentagon’s strategic approach for operationalizing those foundational principles and, more broadly, communicates a framework for how DOD will deliberately leverage AI in a lawful, ethical and accountable manner.

“It is imperative that we establish a trusted ecosystem that not only enhances our military capabilities but also builds confidence with end-users, warfighters, the American public, and international partners. The pathway affirms the department’s commitment to acting as a responsible AI-enabled organization,” Hicks said in a statement shared with FedScoop.

Modern computer systems rely on AI to perform tasks that typically demand at least some human intelligence. Though it’s not new, “technological breakthroughs in the last decade have drastically changed the national security landscape,” Hicks noted in her foreword to the pathway. For that and other reasons, the Pentagon has been increasingly deploying AI in recent years to enable a wide range of functions both on and off the battlefield.

But along with the heaps of benefits the technology can provide, it also holds the potential to introduce high-risk, unintended consequences if not used carefully.

“As the DOD embraces Al, it remains focused on the imperative of harnessing this technology in a manner consistent with our national values, shared democratic ideals, and our military’s steadfast commitment to lawful and ethical behavior,” officials wrote in the new pathway. 

After consulting with leading AI experts for more than a year, the Pentagon in February 2020 officially adopted a series of ethical principles to govern its use of the technology based on the recommendations received. DOD reaffirmed its commitment to the principles in May 2021 and published six foundational tenets that serve as priority areas to guide responsible AI implementation across all its components. 

The tenets include: RAI governance, warfighter trust, AI product and acquisition lifecycle, requirements validation, responsible Al ecosystem, and Al workforce.

This new RAI S&I pathway is founded on and organized around those tenets — and according to Hicks, it “makes [DOD’s] RAI policy tractable for implementation.”

Notably, the pathway also adds new, brief goals to each tenet to more deeply communicate the department’s desired result in each priority area. Under “RAI governance,” for example, the pathway directs officials to modernize governance structures and processes that allow for continuous oversight of DOD’s AI use, and to produce “clear mechanisms” to support users and developers in their RAI implementation as well as provide them with a means to report potential concerns. 

DOD’s release of this pathway comes on the heels of a significant structural shakeup that placed a number of its technology-driving components under a newly established Chief Digital and Artificial Intelligence Office (CDAO). In the document, officials outline the roles of that office and other DOD components as they complete the associated work. 

The CDAO’s recently named RAI Chief Diane Staheli will steer and directly support DOD’s ongoing implementation effort, providing day-to-day expertise to those involved.

“Ultimately, DOD cannot maintain its competitive advantage without transforming itself into an AI-ready and data-centric organization, with RAI as a prominent feature,” officials wrote.