Inspectors General to Jointly Review NSA’s AI Implementation in Signals Operations

The inspectors general of the Department of Defense and the National Security Agency will jointly evaluate the NSA’s integration of artificial intelligence into signals operations.

DOD’s Office of Inspector General said it will review NSA’s efforts against AI guidance issued by the Pentagon and the Intelligence Community.

In a memorandum, DOD OIG said it may revise the evaluation process as needed or based on the suggestions of DOD and NSA management.

The memo was delivered to the director of national intelligence, undersecretary of defense for intelligence and security, DOD chief information officer and the NSA director, Nextgov reported.

Nextgov said the criteria will likely be based on the DOD’s Ethical Principles for AI guidance.

In a May memo about the principles, Deputy Secretary of Defense Kathleen Hicks stressed the importance of existing legal, safety and policy frameworks in AI development.

She said the principles are collectively geared toward creating technically competent AI technologies while maximizing human control and minimizing unintended consequences, including bias.

The DOD’s Joint Artificial Intelligence Center previously said that ethics is a major consideration in developing trustworthy AI.

Alka Patel, head of AI ethics at JAIC and a speaker at a past Potomac Officers Club event, said AI trustworthiness is important for enabling autonomy, decision-making and system execution at fast speeds.

In July 2020, the Intelligence Community also released a set of principles designed to guide intelligence agencies’ ethical development of AI.

Then-Director of National Intelligence John Ratcliffe said the principles and their accompanying framework is significant for data scientists who are using AI to solve intelligence problems.