DoD unveils how it will keep AI in check with ethics principles

The military formally announced serious precautions it will take in the future as it continues to research and invest in artificial intelligence, a technology the Pentagon thinks will be pivotal to worldwide strength.

“The stakes for AI adoption are high,” Air Force Lt. Gen. Jack Shanahan, director of the Joint Artificial Intelligence Center, said Monday at the Pentagon. “AI is a powerful emerging and enabling technology that is rapidly transforming culture, society and, eventually, even warfighting. Whether it does so in a positive or negative way depends on our approach to adoption and use. The complexity and speed of warfare will change as we build an AI-ready force of the future. We owe it to the American people and our men and women in uniform to adopt AI principles that reflect our nation’s values.”

Defense Secretary Mark Esper officially signed off on a five-point AI ethics memorandum that will go into everything from the development of the technology, to the data used to how the technology is implemented.

The principle areas are based on recommendations from a 15-month study by the Defense Innovation Board — a panel of science and technology experts from industry and academia.

The principles are as follows:

  1. Responsible: DoD personnel will exercise appropriate levels of judgment and care, while remaining responsible for the development, deployment, and use of AI capabilities.
  2. Equitable: The Department will take deliberate steps to minimize unintended bias in AI capabilities.
  3. Traceable: The Department’s AI capabilities will be developed and deployed such that relevant personnel possess an appropriate understanding of the technology, development processes, and operational methods applicable to AI capabilities, including with transparent and auditable methodologies, data sources, and design procedure and documentation.
  4. Reliable: The Department’s AI capabilities will have explicit, well-defined uses, and the safety, security, and effectiveness of such capabilities will be subject to testing and assurance within those defined uses across their entire life-cycles.
  5. Governable: The Department will design and engineer AI capabilities to fulfill their intended functions while possessing the ability to detect and avoid unintended consequences, and the ability to disengage or deactivate deployed systems that demonstrate unintended behavior.

What that means for the development and use of AI within the department remains to be seen.  Shanahan and DoD Chief Information Officer Dana Deasy said the way the military responds to these principles will evolve along with the use and development of AI. Deasy said there is no such thing as an end state, rather DoD will continue to learn.

However, certain aspects of DoD’s intent with the principles are clear from the beginning.

“It’s about understanding where these ethics principles need applied,” Shanahan said. “It’s going to be everything from where’s your data come from? What’s the data look like? Is the data representative of a very small sample size instead of a very diverse set of data that would be necessary to develop a robust algorithm. It goes all the way to test and evaluation. What do we need to do to show we are meeting the reliable piece of the five principles?”

Read more: Defense News

Shanahan said when the department fields a capability it will be representative of the principles.

“This will bring even more structure and discipline for how we go about building solutions, fielding solutions and the ongoing operations of those solutions,” Deasy said.

The ethics will also hold people in the AI pipeline accountable. For example, if a developer does not use data that is equitable, then DoD knows where something went wrong to hold that person accountable.

DoD is implementing an AI executive steering committee that will oversee the principles and decide when human intervention is needed in the technology loop.

Copyright © 2020 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.