The tense dispute between the Pentagon and AI firm Anthropic centers on whether the U.S. military may use Anthropic’s Claude model for “all lawful purposes” or must accept explicit limits on applications such as mass surveillance and fully autonomous weapons. The Defense Department set a deadline for Anthropic to agree to broader use or risk losing a lucrative contract and being labeled a supply chain risk, while Anthropic insists on safeguards that bar its technology from enabling wide‑scale monitoring of Americans and AI‑driven targeting without human control. Pentagon chief technology officer Emil Michael argues the military has “made some very good concessions,” promising to put in writing that existing laws and Pentagon policies already prohibit domestic surveillance abuses and autonomous weapons, and inviting Anthropic onto its AI ethics board. However, Anthropic responds that new contract language shows “virtually no progress” and embeds “legalese” that could let the government ignore safeguards at will.
CEO Dario Amodei maintains that the company “cannot in good conscience accede” to Pentagon demands, warning that current frontier AI systems are not reliable enough to operate fully autonomous weapons or to replicate the judgment of “highly trained, professional troops.” He also highlights the risk that AI could assemble “scattered, individually innocuous data into a comprehensive picture of any person’s life,” enabling intrusive surveillance. The Pentagon, backed by the Trump administration’s skepticism toward stringent AI regulation and “woke” models, counters that it must remain prepared for adversaries like China and cannot “say that we’re not going to be able to defend ourselves in writing to a company.” Officials signal they are ready to sever ties, shift to other AI providers, and even consider invoking the Defense Production Act, underscoring a wider ideological clash over how far private firms should shape military rules for emerging AI technologies.
Reference
Jacobs, J., & Walsh, J. (2026, February 26). Pentagon official lashes out at Anthropic as talks break down: “You have to trust your military to do the right thing”. CBS News. https://www.cbsnews.com/news/pentagon-anthropic-feud-ai-military-says-it-made-compromises/