An Ethical Basis for Autonomous System Deployment
Abstract
This project investigated and implemented an ethical basis for deployment of lethality in autonomous robotic systems. Two main thrusts were explored. The first addresses the ethical dimensions of robotic weaponry in two contexts: the robot as an extension of the warfighter and the robot as an autonomous agent acting on behalf of the warfighter. A formal survey has been completed among a broad population of relevant parties including military personnel, the public, policymakers, and roboticists. The results characterize the decision-making space for the deployment of intelligent robotic weaponry, whereby the military can judiciously determine its most effective and appropriate usage. The second component involved the generation of an artificial "conscience" for an intelligent autonomous robotic agent, which applies limits and constraints on its actions as required by the bounds of ethical decision making. These limits are generated from the Laws of War, rules of engagement, and other requirements. The intent is to yield robots that can perhaps act more humanely than humans do under highly stressful conditions; provide warnings in the field to military decision-makers about the potential ethical consequences of tactical actions regarding the use of this technology; and to ensure that accountability is engineered into these systems from the onset.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 24, 2009
- Accession Number
- ADA508646
Entities
People
- Ronald C. Arkin
Organizations
- Georgia Tech Research Corporation