Defense officials want Trump to be ‘very careful’ about use of AI in military

The Trump administration is moving at full speed to integrate artificial intelligence into the U.S. military, sparking an intense national conversation about how to balance rapid technological dominance with essential safety guardrails. While the Pentagon eyes AI as a critical advantage, some military leaders are urging a more measured approach to its deployment.

Commanders Urge Caution

Adm. Frank Bradley, the head of U.S. Special Operations Command, addressed these concerns during a recent special forces conference in Tampa, Florida. He emphasized that the military must be extremely careful when integrating AI into lethal operations. Bradley noted that while the future could involve AI identifying potential targets, humans must remain the final authority to ensure that force is applied exactly as intended.

For some, the focus for AI isn't about pulling the trigger, but about managing the complex data environments troops navigate daily. Officials like Sgt. Maj. Andrew Krogman and acquisition lead Melissa Johnson noted that AI can streamline administrative duties and reduce cognitive workloads, allowing operators to focus on their primary missions rather than replacing human judgment.

Hegseth Pushes for AI Integration

Despite calls for caution, Defense Secretary Pete Hegseth is moving aggressively to evolve the military's tech capabilities. Hegseth has publicly stated his desire for systems that operate without ideological constraints, warning that he would reject any AI models that restrict necessary military applications. President Donald Trump has mirrored this sentiment, recently shelving an AI executive order out of concern that regulation could weaken the United States' lead over global competitors like China.

This push for speed has led to high-profile friction with the private sector. The Pentagon is currently locked in a bitter dispute with Anthropic, the San Francisco-based AI firm. After Anthropic expressed concerns regarding autonomous weapons and mass surveillance, the defense department canceled the company's $200 million contract and designated it a supply chain risk. Anthropic has since filed a lawsuit, while the Pentagon has pivoted toward collaborations with firms like Google, OpenAI, and SpaceX.

Efficiency at Scale

The military's practical application of AI is already visible. Lt. Gen. Michael Conley of the Air Force Special Operations Command noted that his units utilized AI bots to rapidly reclassify top-secret intelligence for drone operators during the Iran war. Furthermore, research from Georgetown University’s Center for Security and Emerging Technology highlights how the Army’s 18th Airborne Corps successfully utilized AI to manage artillery strikes with greater efficiency and fewer personnel.

Helen Toner, the center's interim executive director, suggests that these varying perspectives are all part of a broader, realistic strategy. While the public often underestimates the military's internal caution, commanders are balancing the need for lethal scale with the absolute necessity of avoiding errors like civilian casualties or friendly fire.