The Pentagon has escalated its dispute with artificial intelligence startup Anthropic PBC, threatening to effectively ban the company from the US military’s supply chain. Defence Department spokesman Sean Parnell announced a deadline of Friday 5.01pm in Washington (Saturday 9.01am AEDT) for Anthropic to grant unfettered access to its Claude Gov AI tools, following Anthropic’s insistence on certain safeguards. Anthropic develops AI systems, including the Claude AI assistant, focusing on safety and research. Their work aims to ensure AI benefits humanity.
Parnell stated the unrestricted access is a “simple, common-sense request” to prevent Anthropic from jeopardising critical military operations. A senior Pentagon official confirmed that the Defence Department sent its final offer to Anthropic on Wednesday.
Anthropic has requested that US officials refrain from using its products to create weapons that autonomously target enemy combatants or conduct mass surveillance. The Pentagon has rejected these restrictions, demanding the ability to use Claude, one of the few AI tools approved for classified cloud work, without limitations.
The Defence Department has also threatened to invoke the Defence Production Act to gain access to Anthropic’s software. Parnell’s recent statement marks the department’s first explicit articulation of potential consequences if Anthropic does not comply.
