The U.S. Army has announced two new initiatives, “Break AI” and “Counter AI,” as part of a 500-day plan to integrate artificial intelligence (AI) into its operations while ensuring the safety and security of its troops.
The Army’s acquisition, logistics, and technology (ALT) office is focusing on testing and evaluating AI technologies to ensure their reliability in the field.
The Army recognizes the need to address potential risks associated with AI, including poisoned datasets, adversarial attacks, and trojans.
Young Bang, principal deputy to the assistant secretary of the Army’s ALT, emphasized the importance of developing AI in controlled environments and collaborating with commercial vendors to leverage existing algorithms.
The Army completed a 100-day sprint to explore AI integration into its acquisition processes, examining ways to develop and secure AI technologies.
The “Break AI” initiative focuses on testing and evaluating AI, particularly in the context of artificial general intelligence (AGI), which aims to develop software that matches or surpasses human cognitive abilities.
The Army recognizes the need to test and evaluate AI that can make unpredictable decisions and learn, particularly as it moves towards AGI.
The “Counter AI” initiative aims to protect the Army’s platforms, algorithms, and capabilities from adversarial AI attacks.
Jennifer Swanson, deputy assistant secretary of the Army’s office of Data, Engineering, and Software, highlighted the growing investment in AI by adversaries and the importance of countering potential threats.
While specific details of the Army’s AI initiatives remain classified due to operational security considerations, Swanson stated that information will be shared as the program progresses.