Pentagon Seeks Broader Access to AI Tools on Classified Military Networks

The Pentagon is accelerating efforts to deploy advanced artificial intelligence tools across both unclassified and classified military networks, signaling a major shift in how the U.S. Department of Defense plans to integrate AI into national security operations. According to officials familiar with the discussions, the military wants leading AI companies such as OpenAI, Anthropic, Google, and xAI to make their generative AI models available with fewer usage restrictions, particularly in sensitive defense environments.

During a recent White House event, Pentagon Chief Technology Officer Emil Michael confirmed that the Department of Defense is “moving to deploy frontier AI capabilities across all classification levels.” The initiative reflects ongoing negotiations between the Pentagon and top AI developers over how artificial intelligence will be used in future battlefields increasingly shaped by autonomous drone swarms, robotics, and cyber warfare.

Currently, most AI tools developed for military use operate on unclassified networks primarily used for administrative tasks. OpenAI recently reached an agreement allowing its tools, including ChatGPT, to be used on the Defense Department’s unclassified genai.mil network, which serves more than 3 million personnel. While some safeguards remain, OpenAI has eased several standard user restrictions under this deal. Any expansion into classified systems would require a separate agreement.

Anthropic’s AI assistant, Claude, is already accessible in classified environments through third-party providers, though government users must still comply with the company’s usage policies. Discussions between Anthropic and defense officials have reportedly been more complex, particularly regarding concerns over autonomous weapons targeting and domestic surveillance.

Military leaders believe AI can significantly enhance decision-making by rapidly synthesizing intelligence data. However, experts warn that generative AI systems can produce inaccuracies or fabricated information, posing serious risks in high-stakes classified settings. As debates continue over AI safeguards and military autonomy, the Pentagon’s push highlights the growing role of artificial intelligence in U.S. defense strategy and national security operations.