The U.S. Department of Defense has officially restricted the use of artificial intelligence tools provided by the company Anthropic. According to reports, the Pentagon designated the firm as a "supply chain risk" in February. This decision followed a disagreement between the military and the AI developer regarding the configuration of the software.

The core of the dispute centers on the Pentagon's request for Anthropic to remove or adjust specific safety guardrails embedded within its AI models. Anthropic reportedly refused to comply with these requests, maintaining that its safety measures are essential to the tool's operation. Consequently, the Department of Defense concluded that the software did not meet the necessary security requirements for military integration, leading to the current ban.

While the Pentagon has not provided a detailed public breakdown of the specific risks identified, the "supply chain risk" designation suggests concerns over the software's provenance, data handling, or potential vulnerabilities that could be exploited by external actors. Anthropic, known for its focus on AI safety and constitutional AI, has not publicly detailed the specific guardrails the Pentagon sought to disable. The move highlights the ongoing tension between the U.S. military's desire to adopt cutting-edge commercial AI and the strict security and control standards required for defense operations.