Amidst the prevailing apprehension and uncertainty regarding the future implications of artificial intelligence in weaponry, the Pentagon has recently revamped a policy that is a decade old to dispel any notions of clandestine activities involving “killer robots in the basement,” as stated by a senior official.
The Department of Defense has introduced a relatively new directive that enforces a more stringent review process for the development and authorization of most autonomous weapons, representing a step towards “good governance,” as highlighted by the official. The surge of AI technologies in the military domain, not only in the United States but also in countries like China, has spurred international calls for monitoring the utilization and capabilities of such technologies.
In the past year, the DoD revised its autonomy directive for weapons systems, originally promulgated in 2012, in response to the “dramatic advances in technology happening all around us,” articulated by Deputy Secretary of Defense Dr. Kathleen Hicks, reflecting the significant and rapid evolution of AI over the past decade.
A pivotal alteration in the DoD directive on AI involves the implementation of a senior review policy mandating that any “autonomous weapon systems, including weapon systems with both autonomous and semi-autonomous modes of operation,” must receive approval from a cadre of senior defense officials before the commencement of development and once more prior to fielding any technology.
The revision of the directive introduces stricter regulations in alignment with the cautious approach adopted by the DoD towards AI integration in weapons systems. Furthermore, according to Deputy Assistant Secretary of Defense for Force Development and Emerging Capabilities Dr. Michael C. Horowitz, it serves to elucidate any ambiguities surrounding the Pentagon’s stance on AI and its military applications.
At a recent panel discussion at the Center for Strategic and International Studies (CSIS) Wadhwani Center for AI and Advanced Technologies, Horowitz emphasized that the primary objective behind revising the directive was primarily centered on enhancing governance practices.
The updated directive aims to clarify the permissibility and restrictions concerning the DoD’s policy on AI weapon development. While there may not be any “killer robots in the basement,” the military is actively exploring the potential of AI in various applications.
