HALO Act of 2026
Sponsored By: Senator Schiff, Adam B. [D-CA]
Introduced
Summary
Human authority over lethal decisions would be the central requirement of the Human Authority in Lethal Operations Act of 2026. The bill would create a comprehensive governance framework for autonomous and semi-autonomous weapon systems and related artificial intelligence covering design, safety, testing, oversight, transparency, and alignment with the Law of Armed Conflict and applicable treaties.
Show full summary
- Military commanders and operators would face new design and operational controls that build in human judgment, mandate the ability to reject or challenge AI recommendations, name a designated commander for systems, and require autonomy algorithms that can be rapidly reprogrammed.
- People in conflict zones would benefit from stricter testing and safety standards because systems must demonstrate robustness against adaptive adversaries and safe operation around civilians, with safeguards for densely populated areas to reduce unintended engagements.
- DoD testers, acquisition teams, and oversight bodies would get tougher verification and validation rules, a Chief Digital and Artificial Intelligence Officer role, stronger authority and standards for the Director of Operational Test and Evaluation, a centralized incident reporting repository, advance reviews of test plans, and reporting and whistleblower protections to specified congressional committees.
Your PRIA Score
Personalized for You
How does this bill affect your finances?
Sign up for a PRIA Policy Scan to see your personalized alignment score for this bill and every other piece of legislation we track. We analyze your financial profile against policy provisions to show you exactly what matters to your wallet.
Bill Overview
Analyzed Economic Effects
7 provisions identified: 6 benefits, 0 costs, 1 mixed.
Human control and pre-deployment review
If enacted, the bill would require senior DoD officials to review covered autonomous and semi‑autonomous systems before formal development and again before fielding. Reviews would have to confirm a designated commander can exercise final discretion and that systems limit targets, geography, and times without human approval. The bill would require re‑review when algorithms, mission sets, environments, or targets materially change. Some narrow exceptions would apply and the rule would take effect 180 days after enactment.
Limits on DoD AI uses
If enacted, the bill would prohibit the Department from using covered AI in ways that violate the Constitution, federal law, the Law of Armed Conflict, or international obligations. It would ban profiling people for exercising rights, inferring emotions, inferring protected traits, and tracking people in the United States using commercial data without a court order. The bans would also apply to other federal agencies that receive shared systems or data. The rule would take effect 180 days after enactment.
Stronger testing and fixes for systems
If enacted, the bill would require rigorous verification and validation and realistic developmental and operational tests for every covered system. It would require post‑fielding monitoring, quarterly cyber testing, and a centralized incident repository run by the Director of Operational Test and Evaluation. Within 180 days of the rule taking effect, the Secretary would review all systems in use and must stop using any system that cannot be attested as compliant and send Congress a remediation plan.
New Chief AI officer at DoD
If enacted, the bill would create a Chief Digital and Artificial Intelligence Officer to monitor AI and cybersecurity for covered systems and to advise the Secretary. The officer would set testability standards, issue guidance on red‑team and evaluation practices, and coordinate common tools and benchmarks for verification and validation. The role would take effect 180 days after enactment.
Definitions for covered weapon and AI systems
If enacted, the bill would define terms that decide which systems and people the rules cover. Definitions include "autonomous weapon system," "semi‑autonomous weapon system," "covered artificial intelligence capability," "designated commander," and "unintended engagement." Those definitions would apply across the Act and take effect 180 days after enactment.
Whistleblower rules for DoD AI workers
If enacted, the bill would require the Secretary to update whistleblower protections and procedures for personnel who develop, test, deploy, operate, or use AI. The updates would allow anonymous or confidential reports when appropriate and require investigations and corrective action. Confidentiality would not cover significant misconduct or cases where law forbids it.
Public posting of DoD AI principles
If enacted, the bill would require the Secretary to make the Department's Ethical Principles for Artificial Intelligence and its Responsible AI Strategy publicly available. Any revision would have to be posted at least 30 days before it goes into effect. The rule would take effect 180 days after enactment.
Sponsors & CoSponsors
Sponsor
Schiff, Adam B. [D-CA]
CA • D
Cosponsors
There are no cosponsors for this bill.
Roll Call Votes
No roll call votes available for this bill.
View on Congress.gov