The Pentagon is accelerating efforts to embed artificial intelligence across the U.S. military, even as senior commanders and industry raise alarms about safety and control. Adm. Frank Bradley, who leads U.S. Special Operations Command, cautioned that AI-enabled targeting must preserve human judgment and avoid unintended harm, contrasting with Defense Secretary Pete Hegseth’s push to use AI “any legal way” without ideological limits. The split view comes amid a high-profile clash with Anthropic, which lost a $200 million defense contract after the Pentagon labeled it a “supply chain risk”; the company has sued, alleging retaliation over safety concerns. While commanders say AI can streamline tasks and accelerate decision-making—from downgrading intelligence for field use to identifying targets—policymakers face a balancing act between speed, ethical safeguards and dependency on a handful of tech suppliers.
Related articles:
Project Maven
Anthropic (company)
Unmanned aerial vehicles in the Russian invasion of Ukraine
