NDAA would mandate new DOD steering committee on artificial general intelligence

A version of the annual defense policy bill unveiled by lawmakers Sunday would require Pentagon leadership to form a new panel to look at the military implications of artificial general intelligence and other advanced AI capabilities.

The compromise version of the fiscal 2026 National Defense Authorization Act hammered out by House and Senate conferees includes a provision that would mandate that the secretary of defense establish an “Artificial Intelligence Futures Steering Committee” comprised of top brass, no later than April 1, 2026.

The steering committee would be co-chaired by the deputy secretary of defense and vice chairman of the Joint Chiefs of Staff. Other members would include the service vice chiefs; vice chief of the National Guard Bureau; the undersecretaries of defense for acquisition and sustainment, research and engineering, intelligence and security, and comptroller/chief financial officer; the Pentagon’s chief digital and AI officer; and other officials that the SecDef deems appropriate.

The panel would be tasked with “formulating a proactive policy for the evaluation, adoption, governance, and risk mitigation of advanced artificial intelligence systems by the Department of Defense that are more advanced than any existing advanced artificial intelligence systems, including advanced artificial intelligence systems that approach or achieve artificial general intelligence,” the bill text states.

According to a description from IBM, artificial general intelligence, or AGI, is “a hypothetical stage in the development of machine learning (ML) in which an artificial intelligence (AI) system can match or exceed the cognitive abilities of human beings across any task.”

The Pentagon’s AI futures steering group would also be required to analyze emerging AI models and enabling technologies that could potentially lead to AGI, including frontier and world models, agentic algorithms, neuromorphic computing, cognitive science applications for the development of algorithms or models, infrastructure required to support the deployment of new AI systems at scale, and microelectronics designs or architectures.

Additionally, the committee would be directed to examine the potential “operational effects” of integrating advanced or “general purpose” AI into DOD networks and systems from a technical, doctrinal, training and resourcing perspective, and assess how the tech may affect future military commanders.

Development of “a risk-informed strategy” for the Pentagon’s adoption of the technology would need to include an assessment of “potential effects on commanders of operational commands, including effects related to maintaining oversight of mission command when using artificial intelligence and the capability for humans to override artificial intelligence through technical, policy, or other operational controls,” per the legislation.

Lawmakers also had U.S. adversaries in mind when they drafted the NDAA.

The new steering committee would be tasked with looking at the technological, operational and doctrinal trajectories of America’s foes with regard to advanced AI, “including any pursuit or development by such adversaries of artificial general intelligence,” according to the bill.

After analyzing the threat landscape, Pentagon brass would be tasked with developing options and counter-AI strategies that the department could pursue to defend against that type of future technology.

According to the NDAA, the deputy SecDef would be required to submit a report on the steering group’s findings — including its assessment of broad resource requirements for artificial intelligence — to the congressional defense committees no later than Jan. 31, 2027.

Both chambers of Congress must pass the compromise version of the NDAA and have it signed by the president before it becomes law.