Key Takeaways
- The Trump administration is aggressively pursuing AI to maintain U.S. technological superiority, even as it faces calls for responsible safeguards.
- Senior military leaders, including Adm. Frank Bradley, stress that AI must enhance— not replace— human judgment in lethal operations.
- Defense Secretary Pete Hegseth advocates for unrestricted AI use, rejecting models that “won’t allow you to fight wars.”
- Within the military, AI serves two parallel roles: speeding target identification and strike efficiency, and reducing administrative workload for operators.
- Studies show AI can achieve targeting results comparable to elite units while using far fewer personnel, but human operators remain the final decision‑makers.
- A public clash with AI firm Anthropic over safety concerns has led the Pentagon to label the company a supply‑chain risk and shift contracts to rivals such as Google, OpenAI, and SpaceX.
- Experts warn that without proper guardrails, AI‑enabled warfare risks unintended harms like civilian casualties or friendly fire, even as commanders seek mission success.
Administration’s AI Push and National‑Security Competition
The Trump administration has framed artificial intelligence as a cornerstone of American military advantage, insisting that the United States must stay ahead of rivals such as China. President Donald Trump abruptly canceled a planned AI executive order, warning that any measure that “could dull America’s edge on AI technology” would be unacceptable. He told reporters, “We’re leading China, we’re leading everybody, and I don’t want to do anything that’s going to get in the way of that lead.” This stance places the Pentagon under pressure to adopt AI rapidly while simultaneously addressing calls for responsible safeguards from tech firms and uniformed leaders.
Admiral Bradley’s Call for Caution in Lethal AI
Adm. Frank Bradley, head of U.S. Special Operations Command, warned conference attendees in Tampa that troops “have to be very careful about how we come to (AI’s) employment and its inspiration into the delivery of lethality.” He acknowledged a future where AI could autonomously select targets but emphasized that “we, as humans, have to have the confidence that … it’s going to deliver violence only where we intend it to be delivered.” Bradley’s remarks underscore the senior leadership’s belief that AI must be a tool that supports, not supplants, human moral and tactical judgment.
Secretary Hegseth’s Unconstrained Vision for Military AI
Defense Secretary Pete Hegseth has taken a more permissive line, asserting that the Pentagon should be allowed to use AI “any legal way it sees fit.” Speaking to SpaceX employees in January, he declared he would reject any AI models “that won’t allow you to fight wars” and described his ideal as systems operating “without ideological constraints that limit lawful military applications.” Hegseth’s stance has sparked friction with companies wary of insufficient safety measures, setting the stage for a broader debate over how much latitude the military should have in deploying AI.
Two Parallel AI Worlds Within the Force
Inside the military, AI is being pursued along two complementary tracks. A Pentagon official, speaking anonymously, said efforts focus on creating “functional battlefield tools” that help troops “come up with and identify targets more quickly” and thereby “speed up strikes on those targets.” Meanwhile, officials at U.S. Special Operations Command describe AI as a means to free operators from routine duties. Sgt. Maj. Andrew Krogman, the command’s top enlisted leader, said he sees AI handling administrative tasks to “free up operators or helping modernize how the command does business.” Melissa Johnson, the top acquisition official, added that AI should be “reducing the cognitive workload on mundane tasks,” insisting that “we’re leveraging AI more and more, but it’s not to replace operator judgment, it’s to enhance it.”
Academic Insight: Dual Uses and Proven Efficiency
Helen Toner, interim executive director of Georgetown University’s Center for Security and Emerging Technology, observed that both descriptions are accurate. “There are a huge number of potential uses for AI in these kinds of bureaucratic settings, which the U.S. military is actively exploring,” she said. Toner pointed to a case study from her center showing that the Army’s 18th Airborne Corps used AI to direct artillery strikes “just as efficiently as the best unit in recent American history” while employing 2,000 fewer service members. She concluded that “Human operators are still the ones making crucial decisions, but AI … is making it possible to operate with a new level of speed and scale.”
AI’s Role in Accelerating Targeting While Keeping Humans in the Loop
The same Toner analysis highlights that AI’s primary battlefield value lies in accelerating the kill chain without removing human oversight. By rapidly processing intelligence and suggesting targets, AI enables commanders to act faster, yet the final authority to release weapons remains with human operators. This balance aims to preserve accountability and reduce the risk of errors such as misidentification or unintended collateral damage, even as the technology expands the scale and tempo of operations.
Public Clash with Anthropic Over Safety and Ethics
The integration of AI into defense has spilled into the public arena through a heated dispute with AI firm Anthropic. After CEO Dario Amodei resisted Pentagon pressure over concerns about how its chatbot Claude might be used in classified networks, both President Trump and Secretary Hegseth accused Anthropic of “endangering national security.” The Pentagon responded by labeling the San Francisco‑based company a supply‑chain risk, terminating its $200 million defense contract and barring other government contractors from collaborating with it. Anthropic sued, alleging illegal retaliation and claiming the designation was meant to guard against foreign sabotage, not to punish a domestic vendor. The Defense Department has since pivoted to rivals such as Google, OpenAI, and SpaceX, seeking AI that can “augment warfighter decision‑making in complex operational environments.”
Toner’s Warning: Balancing Lethal Effect with Unintended Harm
Reiterating the need for caution, Toner warned that the public often underestimates how carefully the military evaluates new technologies. She stressed that commanders’ objectives are twofold: achieving lethal effects at scale while avoiding unintended consequences like friendly fire, civilian casualties, or erroneous target identification. Her commentary captures the central tension driving the current debate—how to harness AI’s speed and analytical power without eroding the ethical and legal norms that govern the use of force.
Looking Ahead: Innovation Versus Guardrails
As the Trump administration continues to champion AI as a strategic asset, the military finds itself navigating a fine line. Senior leaders like Bradley and Toner advocate for robust safeguards, insisting that AI must remain subordinate to human judgment, especially in lethal contexts. In contrast, figures such as Hegseth push for rapid, unencumbered adoption to preserve America’s technological edge. The outcome of this tug‑of‑will—shaped by contractual battles, public scrutiny, and evolving doctrine—will determine whether AI becomes a force multiplier that enhances precision and restraint, or a source of new risks that challenge the very principles of lawful warfare.
https://fortune.com/2026/05/31/special-operations-commander-frank-bradley-ai-targets-pete-hegseth-anthropic/