Army Introduces New AI Strategies for Troop Safety

The U.S. Army has unveiled new initiatives aimed at enhancing the safety and effectiveness of artificial intelligence (AI) within its operations. Announced this week, the "Break AI" and "Counter AI" programs are part of a broader 500-day plan to ensure reliable AI deployment in the field while protecting against potential adversarial uses of AI technology. The Army's acquisition, logistics, and technology (ALT) office is spearheading these efforts, focusing on both the implementation and development of AI in collaboration with external partners. During a recent tech conference in Georgia, Young Bang, principal deputy to the assistant secretary of the Army’s ALT, emphasized the importance of addressing challenges such as poisoned datasets and adversarial attacks. He noted that developing AI in a controlled environment owned by the Department of Defense would facilitate safer integration of third-party algorithms into Army programs, thereby avoiding competition with commercial vendors. The "Break AI" initiative will explore advancements in artificial general intelligence (AGI), which aims to create software capable of matching or exceeding human cognitive abilities. This initiative underscores the Army's commitment to not only advancing AI technology but also ensuring robust defenses against its potential misuse. Bang highlighted the complexities of testing AI systems, particularly as they evolve toward AGI, where outcomes and behaviors may be unpredictable. As the Army moves forward, it aims to secure its platforms and algorithms from threats while simultaneously countering adversarial capabilities, marking a significant step in the military's approach to AI technology.