Pentagon Seeks Ethical Guidelines For Use Of AI In Warfare

The U.S. Dept. of Defense is seeking to develop new ethical guidelines for the use of artificial intelligence in warfare, reports Defense One.

Pentagon officials have asked the Defense Innovation Board (DIB) to draft new guidelines on artificial intelligence to help the department “safely and responsibly” employ new technologies, said Josh Marcuse, the executive director of the board.

The new guidelines will replace a 2012 policy, which says only that a human being must have veto power over any action an autonomous system might take in combat. The four-page document fails to address emerging issues, such as the uses of AI for decision support and predictive analytics.

Pentagon leaders will determine how to put the principles into practice. The goal is for a range of military leaders to find them useful for preparing new strategies, operational concepts and training guides.

The DIB, chaired by Silicon Valley corporate leaders, will begin the process on Jan. 23 with a public listening session at the Belfer Center for Science and International Affairs at Harvard University. 

The board hopes to solicit comments from academics and ethicists who have previously expressed concern over the military’s use of AI. One common concern is data bias or the occurrence of algorithmic decisions based on inaccurate information. Such errors could potentially result in unintended deaths.

A list of guideline recommendations is expected to be made public in June.