Establishes Department of Defense policy on the use of autonomous and semi-autonomous weapons systems, including testing & evaluation requirements and requirements that humans be able to intervene should such weaponry begin to operate in unintended ways.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding DoD Directive with mandatory requirements enforced through military chain of command, using extensive mandatory language ('shall', 'must', 'will') and establishing formal approval, monitoring, and enforcement mechanisms.
The document has good coverage of approximately 8-10 subdomains, with strong focus on AI system security (2.2), cyberattacks and weapon development (4.2), competitive dynamics (6.4), goal misalignment (7.1), dangerous capabilities (7.2), lack of robustness (7.3), and lack of transparency (7.4). Coverage is concentrated in security, weapons development, and AI safety domains, reflecting the document's focus on autonomous weapon systems governance.
This is an internal DoD policy directive that governs the Department of Defense's own operations in developing and deploying autonomous weapon systems. As such, it primarily governs the National Security sector and Public Administration. The directive does not regulate external commercial sectors but rather establishes internal governance for military AI applications.
The document comprehensively covers all AI lifecycle stages from planning through operational monitoring. It emphasizes verification and validation, testing and evaluation throughout development, and post-deployment monitoring. The directive establishes requirements spanning the entire lifecycle of autonomous weapon systems.
The document explicitly addresses AI systems and AI models, particularly in the context of autonomous weapon systems. It references AI capabilities extensively and establishes AI Ethical Principles. The document does not explicitly mention frontier AI, general purpose AI, task-specific AI, foundation models, generative AI, predictive AI, open-weight models, or specific compute thresholds.
Office of the Under Secretary of Defense for Policy; Department of Defense
The document is issued by the Office of the Under Secretary of Defense for Policy as the Originating Component, and approved by the Deputy Secretary of Defense Kathleen H. Hicks, indicating DoD policy-making authority proposed this directive.
Under Secretary of Defense for Policy (USD(P)); Under Secretary of Defense for Research and Engineering (USD(R&E)); Under Secretary of Defense for Acquisition and Sustainment (USD(A&S)); Vice Chairman of the Joint Chiefs of Staff (VCJCS); Director, Operational Test and Evaluation (DOT&E); General Counsel of the Department of Defense (GC DOD); Chief Digital and Artificial Intelligence Officer (CDAO)
Multiple senior DoD officials are designated with approval, oversight, and enforcement authority over autonomous weapon systems, including mandatory approval requirements before development and fielding.
Director, Operational Test and Evaluation (DOT&E); Under Secretary of Defense for Research and Engineering (USD(R&E)); Chief Digital and Artificial Intelligence Officer (CDAO); Secretaries of the Military Departments; Commander, USSOCOM; Autonomous Weapon System Working Group
The directive establishes comprehensive monitoring responsibilities including post-fielding monitoring, operational test and evaluation, AI capability monitoring, and a dedicated working group to support reviews.
OSD; Military Departments; Office of the Chairman of the Joint Chiefs of Staff; Joint Staff; Combatant Commands; Office of Inspector General of the Department of Defense; Defense Agencies; DoD Field Activities; United States Special Operations Command (USSOCOM)
The directive explicitly applies to all DoD organizational entities involved in the design, development, acquisition, testing, fielding, and employment of autonomous and semi-autonomous weapon systems, making them both developers and deployers of these AI systems.
9 subdomains (6 Good, 3 Minimal)