[Lilux-info] "Ethics Dilemma in Killer Bots"

"Ethics Dilemma in Killer Bots" The Australian (01/16/07) P. 29; Argy, Philip Guard robots being deployed on the northern border of South Korea are capable of firing on human targets without receiving any direct commands from humans, and have brought up many important ethical questions. Each Intelligent Surveillance and Security Guard Robot will be equipped with a daylight camera capable of identifying targets within a 4-kilometer radius, and an infra-red night vision camera that has a range of 2 kilometers. While humans can use a joystick and touchscreen to control the robots, they are programmed to respond autonomously when an intruder does not provide a correct password. The robot's responses include sounding an alarm, using non-lethal force, or firing a machine gun or rifle; these would be the world's first robots with such capabilities. While the manufacturer says the robots are superior to human guards because they are immune to weather conditions and fatigue, many point out that a human soldier could utilize discretion and understand the consequences of his actions. Australian Computer Society's Mike Bowern expresses concerns over the potential for "software and hardware defects" to "influence the robot's conduct." He also points out that little is known of the ethical considerations taken by the robots' designers, or any code they must follow, since Korea doesn't have an independent professional association such as the ACM or the ACS, and the Korean Ministry of Information and Communication seems to place greater importance on technical aspects than it does on professional or ethical concerns. Many worry that these robots could eventually be sold to private customers. Computer ethicist James Moor points out the robots could not be held legally or morally responsible for their actions, leaving such responsibility up to technology professionals. <http://australianit.news.com.au/articles/0,7204,21064361%5E15309%5E%5Enbv%5E,00.html>