Accession Number:

AD1083268

Title:

AI Safety: An Action Plan for the Navy

Descriptive Note:

Technical Report

Corporate Author:

Center for Naval Analyses Alexandria United States

Personal Author(s):

Report Date:

2019-09-01

Pagination or Media Count:

52.0

Abstract:

In light of the Navys stated commitment to using AI, and given the strategic importance of AI safety, we provide the Navy with a first step towards a comprehensive approach to safety. We use a risk management approach to frame our treatment of AI safety risks identifying risks, analyzing them, and suggesting concrete actions for the Navy to begin addressing them. The first type of safety risk, being technical in nature, will require a collaborative effort with industry and academia to address. The second type of risk, associated with specific military missions, can be addressed in a combination of military experimentation, research, and concept development to find ways to promote effectiveness along with safety. For each types of risk, we use examples to show concrete ways of managing and reducing the risk of AI applications. We then discuss institutional changes that would help promote safety in the Navys AI efforts.

Subject Categories:

  • Cybernetics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE