THE PROPOSED MULTI-AGENT REINFORCEMENT LEARNING ALGORITHM AGAINST SPY AGENTS WILL LEAD TO A NEW RESEARCH AREA REGARDING THE SCIENTIFIC AND TECHNICAL ROBUSTNESS AS WELL AS THE ETHICS AND SAFETY OF AI FOR LARGE-SCALE UAV TEAMS. THIS PROPOSAL WILL ALSO CONTRIBUTE TO THE SCIENTIFIC COMMUNITY IN THE LONG TERM BY BRIDGING THE GAP BETWEEN MULTI-AGENT DECISION-MAKING THEORY AND REAL-WORLD APPLICATIONS THAT RELY HEAVILY ON INTERCONNECTED COMPLEX SYSTEMS SUCH AS POWER SYSTEMS, ROBOTICS, ETC. THE LIMITATIONS AND PHYSICAL PERFORMANCE OF THE MULTI-AGENT TEAM IS ALSO EVALUATED IN A NOVEL ASPECT, I.E., THE SPY AGENT ATTACK. ALL DATA AND ALGORITHMS IN THE PROPOSED RESEARCH WILL BE RELEASED FOR PUBLIC USE IN THE FORM OF OPEN-SOURCE SOFTWARE. THE PROPOSED PROJECT WILL PROVIDE OPPORTUNITIES TO TRAIN UNDERGRADUATE AND GRADUATE STUDENTS AT UNIVERSITY OF WYOMING AND UNIVERSITY OF NEVADA. LIVING IN THE LEAST POPULATED STATES IN THE US, WYOMING AND NEVADA RESIDENTS NEED MORE EXPOSURE TO CUTTING-EDGE AI TECHNOLOGIES, AND THIS PROJECT WILL SERVE THAT PURPOSE. SPECIAL ATTENTION WILL BE GIVEN TO THE MINORITIES IN THE WYOMING AND COLORADO AREA THROUGH THE CO-WY ALLIANCE FOR MINORITY PARTICIPATION, AND THE UNIVERSITY OF NEVADA RENO FIRST GENERATION MINORITY STEM PROGRAM. THE PI WILL ALSO PARTICIPATE IN OUTREACH ACTIVITIES TO HELP PROMOTE PUBLIC KNOWLEDGE OF AI TECHNIQUES IN THE STATE OF WYOMING AND NEVADA.