This bill, titled the Human Authority in Lethal Operations Act of 2026 (HALO Act), establishes comprehensive design and safety requirements for autonomous and semi-autonomous weapon systems (AWS/SAWS) used by the Department of Defense. It aims to ensure that human operators maintain ultimate discretion and control over the use of force, even as artificial intelligence capabilities are integrated into military operations. The legislation defines key terms such as "autonomous weapon system" and "semi-autonomous weapon system," emphasizing their varying degrees of autonomous target engagement. A core provision mandates that a designated commander be identified as accountable for each engagement involving these systems, regardless of AI's contribution. The Secretary of Defense must promulgate rules for a clear chain of command, mirroring traditional military operations. Furthermore, system designs must incorporate capabilities and interfaces that require the designated commander to exercise ultimate discretion, judgment, and control over the weapon system's actions, targets, and geographic scope, preventing autonomous expansion or modification. Each system must be designed to operate within designated timeframes, geographic areas, and target sets , consistent with human operator intentions. It must also require independent review before using force against unauthorized targets or expanding operational scope, and terminate engagements if parameters are not met, awaiting further human evaluation. The bill emphasizes the creation and maintenance of detailed records for post-engagement review and mandates consistency with the Department's Ethical Principles for Artificial Intelligence and the Responsible Artificial Intelligence Strategy and Implementation Pathway , which must be publicly available. To ensure transparency and accountability, systems must be designed with technologies and data sources that are auditable and explainable by relevant personnel. Robust system safety, anti-tamper mechanisms, and cybersecurity are also required. Additionally, systems on unmanned platforms must prevent autonomous selection and engagement of targets in the event of degraded or lost communications, ensuring human control is retained. The bill establishes an advanced review and approval authority for autonomous and semi-autonomous weapon systems before formal development and fielding. This process involves joint verification by senior DoD officials, ensuring the system design incorporates human discretion, accounts for risks to civilians, and includes plans for rigorous verification, validation, and comprehensive legal analysis. A separate pre-fielding review confirms that system capabilities, training, and safety measures are adequately implemented and demonstrated. The Chief Digital and Artificial Intelligence Officer is tasked with monitoring and evaluating AI capabilities, formulating testable ethical requirements, and guiding rigorous testing, including adversarial assessments. All systems must undergo rigorous hardware and software verification and validation , along with realistic developmental and operational testing in realistic conditions, assessing human operator response and system robustness. The legislation explicitly prohibits certain uses of covered artificial intelligence capabilities, particularly those that violate civil liberties, such as profiling based on protected characteristics or tracking individuals without judicial authorization, and forbids removing humans from critical decision-making chains, like nuclear weapons employment. Enhanced whistleblower protections are mandated for personnel reporting concerns about AI systems, ensuring anonymity where appropriate and compliance with existing laws. The Secretary must submit semiannual and annual reports to Congress on AI use cases, operational incidents, compliance, and infrastructure needs. The Act takes effect 180 days after enactment, requiring a review of all existing systems and immediate cessation of use for any non-compliant ones, pending remediation or responsible retirement.