A public dispute between the US Department of Defense and Anthropic (maker of the AI model Claude) is revealing how central advanced AI has become to US military operations.
Anthropic refused to remove contractual safeguards that block use of Claude for mass domestic surveillance and fully autonomous weapons. This stance conflicted with Pentagon demands for AI that could be used for “all lawful purposes.”
In response, the White House ordered federal agencies to phase out Anthropic’s Claude from government systems. Despite this, Claude has already been integrated into classified workflows like intelligence analysis and operational planning.
OpenAI subsequently reached a deal with the Pentagon under broader terms, allowing military use of its AI for “all lawful purposes” while including engineering and policy safeguards.
Analysts say the clash highlights tension between vendor-enforced ethical red lines and government desires for frictionless AI deployment in high-consequence missions. How much control private firms should retain over military uses of AI remains a central debate.
Related:
The New Frontlines of Global Conflict
