Recently, the U.S. Army held an artificial intelligence technology drill with strong combat significance, highlighting its strong willingness to promote combat effectiveness through artificial intelligence. As the US military continues to promote the construction of artificial intelligence, other countries will catch up, and the international security pattern may face more uncertainties due to the artificial intelligence arms race. Several US media reported that in a recent exercise, the US Army successfully used artificial intelligence to “accelerate target targeting” without human intervention. It is reported that the exercise code named “2020 fusion” was carried out in the harsh desert environment of Yuma proving ground, aiming to help the US Army learn how to fight in multi domain battlefield and try new technologies. The exercise lasted from August 11 to September 23 and was divided into three stages. In the penetration phase, the US army uses space and air sensors to identify the targets forming the “anti intervention / area denial” defense; in the disintegration phase, the US Army fighters are responsible for clearing the enemy’s long-range precision firepower; in the utilization phase, the US army ground forces defeat their counterparts in the operation. < p > < p > for the role of artificial intelligence technology in the drill, Ross Coffman, head of the cross functional team of the “next generation chariot” of the US Army, believes that there are three main aspects. First, relying on the “storm” artificial intelligence system, it helps to update the shooter’s common operation pictures and instructions used to strike targets; second, it uses ground robots and small UAVs to conduct digital mapping of the battlefield and transmit information; third, it uses artificial intelligence to identify targets. With the advance of the exercise, the US Army has modified the algorithm to better meet the needs of target recognition Please. The high-level of the U.S. Army praised the AI exercise. “Most of the systems used in this exercise are in the experimental phase, and it took us six weeks to mature the technology, which could take two to three years to operate in the laboratory,” said John Murray, commander of the future command of the US Army Kaufman also stressed that the exercise is an important milestone in the actual combat application of the US Army’s artificial intelligence, which has a symbolic significance that cannot be underestimated. < / P > < p > the first is to develop the auxiliary system of combat vehicles. According to the US defense news website, the U.S. Army is installing artificial intelligence systems for combat vehicles to enable them to drive on their own, select targets independently and determine the priority of attacks, so as to reduce the burden of soldiers in multi domain operations. < / P > < p > the second is to assist the anti UAV system. The U.S. Army and Raytheon are accelerating the development and deployment of an artificial intelligence assisted anti UAV system to deal with the threat of small UAVs in close combat, according to the US Defense one website. The anti UAV system upgrades artificial intelligence and machine learning algorithm, and can be combined with command and control system to provide time sensitive target threat data to decision makers. At present, Raytheon is continuing to develop data fusion algorithms and use artificial intelligence technology to analyze the data collected by sensors to optimize the key operational decision support data. The third is to assist the decision-making of battlefield tankers. According to today’s Russian television website, the U.S. Army is developing an all-round artificial intelligence technology to calculate how to dominate tank warfare by analyzing combat data. The project code named “quarterback plan” aims to provide the tank personnel with artificial intelligence assistant, collect all available battlefield data, analyze it and propose the best way to attack opponents. Because AI can collect data from satellites, UAVs, radars, and cameras placed on soldiers, the technology will fundamentally accelerate operational decision-making. In general, the US Army has made great strides in the development and application of artificial intelligence, which is an epitome of the US Army’s great emphasis on the generation of AI combat effectiveness. Recently, the United States has invited more than 100 experts from 13 countries to gather in the United artificial intelligence center to discuss the ethical issues in the application of artificial intelligence. According to outside comments, the move shows that the United States and its allies intend to seize the “moral high ground” of AI applications. Looking forward to the future, artificial intelligence will still be the top priority in the construction and development of the US military. According to John Murray, commander of the U.S. Army’s future command, the future battlefield is “extremely active battlefield”. The battlefield is full of chaotic and rapidly changing complex information, which requires rapid analysis. The results of future war are likely to be determined by the speed of decision-making. Therefore, the US Army must integrate artificial intelligence programs into almost all future systems. < / P > < p > in this context, other countries are bound to accelerate in this AI competition, and constantly bring forth new ideas in the fields of R & D, test, train and equipment, and actual combat, etc., guided by operational effectiveness. As a result, the global AI arms race may become more and more intense, and its impact on the international security pattern deserves high attention.