Humanitarian Algorithms : A Codified Key Safety Switch Protocol for Lethal Autonomy
This addresses safety and legal compliance issues for military and security applications, but it is incremental as it builds on existing regulatory techniques.
The paper tackles the problem of ensuring lethal autonomous weapons comply with International Humanitarian Law by proposing a codified key safety switch protocol, advocating for a machine learning-enabled approach to regulate such systems.
With the deployment of lethal autonomous weapons, there is the requirement that any such platform complies with the precepts of International Humanitarian Law. Humanitarian Algorithms[9: p. 9] ensure that lethal autonomous weapon systems perform military/security operations, within the confines of International Humanitarian Law. Unlike other existing techniques of regulating lethal autonomy this scheme advocates for an approach that enables Machine Learning. Lethal autonomous weapons must be equipped with appropriate fail-safe mechanisms that locks them if they malfunction.