Senator Adam Schiff Introduces Bill to Mandate Human Oversight of AI-Driven Lethal Weapons

On Monday, Sen. Adam Schiff (D-CA) introduced the Human Authority in Lethal Operations Act. The bill would require a designated commander to retain final authority over any use of force involving autonomous or semi-autonomous weapons that rely on artificial intelligence. It also obliges the Department of Defense to keep detailed records of decision-making processes—such as target selection—and to subject all such systems to a formal review. The legislation further bars the Pentagon from employing AI to launch nuclear weapons, to surveil Americans engaged in constitutionally protected activities, and from purchasing personal data on U.S. citizens from third-party vendors.

The HillSen. Schiff introduces bill restricting Defense Department AI use

Schiff’s proposal follows a broader push by Senate Democrats to embed “guardrails” on military AI. Earlier this year, the Pentagon entered a public dispute with the AI firm Anthropic over the terms of use for its language model, highlighting uncertainties about civilian-military AI interfaces. Concurrently, Sen. Kirsten Gillibrand (D-NY) and Sen. Elissa Slotkin (D-MI) have announced parallel bills or amendments to the annual National Defense Authorization Act (NDAA) that seek comparable restrictions on autonomous weapons, nuclear-launch AI, and domestic surveillance.

The HillSen. Schiff introduces bill restricting Defense Department AI use

• Sen. Adam Schiff (D-CA) – Sponsor of the Human Authority in Lethal Operations Act; long-time advocate for congressional oversight of defense technology.

The Hill

• Sen. Kirsten Gillibrand (D-NY) – Introduced a companion bill targeting AI-enabled nuclear launch and surveillance capabilities.

The Hill

• Sen. Elissa Slotkin (D-MI) – Preparing a related amendment to the NDAA focused on AI guardrails.

The Hill

• U.S. Department of Defense – The primary agency whose AI-enabled weapon systems would be subject to the new oversight requirements.

The Hill

Sen. Schiff emphasized that recent developments “show an urgent need for commonsense guardrails” to align AI use with national security and privacy priorities. He argued that while AI can enhance defense capabilities, “we cannot depend on technology alone to guide us, particularly when the risks of harm can be fatal.” Gillibrand’s proposal similarly stresses the necessity of preventing AI from controlling nuclear arsenals or infringing on civil liberties.

The Hill

If enacted, the act would institutionalize a human-in-the-loop requirement, potentially slowing the deployment of fully autonomous strike platforms. It would also create a statutory barrier against AI-driven domestic surveillance, reinforcing constitutional protections. The legislation could set a precedent for future AI governance across other federal agencies.

The Hill

Current coverage provides no documented opposition from defense officials or industry groups, nor does it detail the bill’s timeline for committee consideration. Further clarification is needed on how the record-keeping provisions will be enforced and audited.

The Hill

• “clear chain of command” — *Human Authority in Lethal Operations Act*, legislation

The Hill

• “The past few months have shown us that there is an urgent need for commonsense guardrails to ensure the Defense Department’s use of AI is in line with Americans’ national security and privacy priorities,” — Sen. Adam Schiff, U.S. Senator (D-CA)

The Hill

• “There are good reasons to use AI technology to advance our national security,” — Sen. Adam Schiff, U.S. Senator (D-CA)

Sen. Schiff introduces bill restricting Defense Department AI use

• “we cannot depend on technology alone to guide us, particularly when the risks of harm can be fatal.” — Sen. Adam Schiff, U.S. Senator (D-CA)

Sen. Schiff introduces bill restricting Defense Department AI use

Schiff’s bill is expected to be referred to the Senate Armed Services Committee, where it may be debated alongside Gillibrand’s and Slotkin’s proposals. Lawmakers anticipate filing amendments to the NDAA later this year, and the Pentagon has indicated it will submit a compliance review of existing AI-enabled systems.

The Hill