A coalition of artificial intelligence policy groups is urging Congress to include specific restrictions on lethal autonomous weapons in the annual National Defense Authorization Act (NDAA), a move that could reshape how the U.S. military develops and deploys AI-powered combat systems. The appeal, reported by The Hill on June 3, 2026, targets the most consequential defense spending bill of the year, directly affecting Pentagon acquisition officers, defense contractors, and international allies who rely on U.S. weapons standards.
Who is affected
The primary groups affected are the AI policy organizations themselves, including those focused on arms control and ethical technology deployment. These groups argue that without explicit guardrails, the Department of Defense could accelerate development of weapons that select and engage targets without human intervention. Defense contractors such as Lockheed Martin, Boeing, and emerging AI firms like Palantir and Anduril would face new compliance requirements if the NDAA language is adopted, potentially slowing their development timelines for autonomous drone swarms and missile defense systems.
U.S. military service members are directly affected, as the policy would mandate that a human remain “in the loop” for critical targeting decisions. This contrasts with current trends toward increasing automation in systems like the Air Force’s Collaborative Combat Aircraft and the Navy’s unmanned surface vessels. International partners in NATO and the Five Eyes intelligence alliance watch closely, as U.S. NDAA provisions often become de facto global standards for allied weapons exports.
What changes
The most significant change would be a statutory requirement for meaningful human control over lethal autonomous weapons. Under the proposed guardrails, any weapon system capable of identifying and attacking targets autonomously would need to undergo stricter testing, certification, and reporting to Congress. This represents a shift from the Pentagon’s current directive, DOD Directive 3000.09, which already requires human oversight but does not carry the force of law and can be modified by internal policy updates.
If enacted, the NDAA language would also require the Secretary of Defense to submit an annual report to Congress detailing all development and deployment of autonomous weapons. This transparency measure would affect how the Pentagon budgets for AI programs, potentially diverting funds from rapid prototyping to compliance and documentation. The change would affect the Pentagon’s Joint Artificial Intelligence Center and the newly created Chief Digital and Artificial Intelligence Office, which are currently pushing for faster integration of AI into combat systems.
For industry, the guardrails would create new legal liabilities. Contractors would need to demonstrate that their systems cannot operate without a human making the final decision to use lethal force. This could lead to redesigns of software architectures and testing protocols, adding months to development cycles for systems like the Army’s Lower-Tier Air and Missile Defense Sensor and the Air Force’s Skyborg drone program.
What to watch for next
The key next step is the markup of the NDAA by the House and Senate Armed Services Committees, expected in late June and July 2026. Observers should watch whether the guardrails language survives committee votes, where defense industry lobbyists and pro-automation Pentagon officials are likely to push back. A compromise could emerge that requires human oversight only for certain categories of weapons, such as those capable of mass casualty attacks, while exempting smaller loitering munitions.
Another factor to watch is the position of the Biden administration. While the White House has issued an executive order on AI safety in October 2023, it did not specifically target lethal autonomous weapons. A formal administration stance on the NDAA language would signal how the executive branch views the balance between military innovation and ethical constraints.
Finally, international repercussions will be significant. If the U.S. enacts legal guardrails, it could strengthen ongoing negotiations at the United Nations on a treaty governing autonomous weapons, a process that has stalled for years. Conversely, if the NDAA language is weakened or removed, critics warn it could embolden other nations—particularly China and Russia—to accelerate their own autonomous weapons programs without similar constraints. The outcome of this legislative fight will set a precedent for how the world’s largest military power governs the future of warfare, with consequences extending well beyond the 2026 defense budget cycle.
