Anthropic’s recent court filings clarify that the company cannot manipulate its generative AI model, Claude, once in operation by the US military. Responding to concerns from the Trump administration, Thiyagu Ramasamy, Anthropic’s public sector head, emphasized that the company lacks the capability to alter Claude’s functionality or shut down access during military operations. This assertion comes amid ongoing disputes with the Pentagon regarding the use of Anthropic’s AI for national security. Defense Secretary Pete Hegseth categorized Anthropic as a supply-chain risk, delaying Department of Defense (DoD) use of the software. As a result, various federal agencies are severing ties with Claude. Anthropic is contesting the ban in court, seeking an emergency order for reversal. Ramasamy highlighted that Anthropic cannot access military data or interfere with operational decisions. Meanwhile, the DoD is implementing measures to ensure Anthropic cannot make unilateral changes to existing Claude systems, reinforcing its commitment to maintaining operational integrity.
Source link
