The Pentagon’s AI acceleration hub has begun operational prototyping of custom commercial machine learning models designed to monitor and analyze adversarial media and related data, Defensescoop reported.

The same was revealed by a Defense Department spokesperson shared in response to questions regarding how and why the Chief Digital and AI Office (CDAO) is rapidly prototyping BigBear.ai’s Virtual Anticipation Network (VANE) to advance complex information operations across the military. 

These models aim to enhance U.S. national security missions by providing rapid, high-tech assessments and predictive insights.

The machine learning tools and datasets within VANE could help identify the relationships between datasets over time, to inform senior leader decisions and provide key strategic context to DOD decisions,” the spokesperson said.

Recently, the Defense Information Systems Agency which manages the Pentagon’s IT networks banned the use of Chinese AI Startup DeepSeek’s AI models. 

DeepSeek’s new AI chatbot remained accessible for at least two days before the Pentagon took action to block it, according to a defense official familiar with the matter.

Other government agencies to ban DeepSeek include NASA and the U.S. Navy.