This bill, titled the "AI Guardrails Act of 2026," aims to establish clear limitations on the Department of Defense's (DoD) use of artificial intelligence (AI) while acknowledging the need for aggressive AI adoption. It specifically prohibits the DoD from using AI for three critical functions: the execution of launching or detonating a nuclear weapon , the monitoring, tracking, profiling, or targeting of individuals or groups in the United States without an individualized, articulable legal basis , and the employment of lethal force by autonomous weapon systems without appropriate levels of human judgment and supervision . The bill does provide a mechanism for the Secretary of Defense to waive the prohibition on autonomous lethal force for up to one year, or renew such a waiver, if extraordinary national security circumstances necessitate it. This waiver requires a written certification to congressional defense committees, confirming that the system's probability of error does not exceed that of trained human operators. Furthermore, the Secretary must notify Congress within five days of issuing or modifying a waiver, providing a detailed rationale, system description, operational parameters, and test results, which can include a classified annex.