As the Defense Department continues to embrace artificial intelligence and even plans to possibly use it in operations as soon as next year, the Air Force is solidifying its focus on the technology by releasing a new strategy that addresses the use, ethics and workforce needs of AI in the service.
The strategy, released Thursday, identifies AI as crucial to fielding a future Air Force, executing multi-domain operations and confronting threats before they reach the level of actual combat.
Signed by acting Air Force Secretary Matt Donovan and Chief of Staff Gen. David Goldfein, the document hones in on five focus areas that the service thinks will properly develop AI for the service’s needs.
One focus area identifies data as a strategic asset, given that AI will need mass amounts of data that standardized for use by AI algorithms.
“We will review our corporate processes and reform underlying critical aspects of agility principled on government-purpose data rights to ensure we are consistently generating training quality data for algorithmic development,” the strategy states. “It is a priority to architect solutions that provide the shortest path between development and operational events.”
To do that, the Air Force plans on setting policies and standards for data and creating controlled and searchable data stores to make data more accessible to airmen.
The Air Force already singled out the need to standardize its data and make it more usable for AI in its recent digital strategy. That strategy takes a more macro perspective and explicitly states that the service will use AI to help airmen make decisions faster.
The Air Force already laid the groundwork by restructuring its chief information office and creating a chief data officer position. The service is putting effort into figuring out what to do with all the data it has, how to combine it productively and then use it quickly in a beneficial way.
The AI strategy, however, is explaining how the Air Force will ensure that one item of the digital strategy is accomplished.
The Air Force’s approach to AI will also drive down technological barriers to entry for AI.
“We will take steps to begin building a commercially relevant delivery model, with due regard to supply chain safety, security and trust,” the strategy states. “We will partner with top technological providers, both within and outside the federal government, in a more cost effective manner, scale development operations and prototyping and share lessons learned throughout the total force.”
In that same vein, the Air Force wants to take advantage of the publicly available AI technology to solve issues.
To do that it plans to host and provide shareable solutions enterprise-wise and to create secure operating environments for prototyping algorithms.
Of course, AI has potential to get out of hand. One only needs to look to HAL 9000 in the space epic 2001: A Space Odyssey to see some of the more fantastical variations of AI gone wrong. But in a more primitive sense, it’s possible that AI could breach citizen privacy or make decisions many feel should only be made my humans.
To address the ethical evolution of AI, the Air Force strategy states the service “will commit to staying engaged, informed and accountable.”
The Air Force plans to engage in dialogue with other nations, and top minds on the ethical, moral and legal implications of deploying AI in the military.
Finally, the strategy addresses the workforce issues AI will create. DoD is having trouble recruiting and retaining technological talent. The Air Force strategy states the service will design practical policies and incentives that foster talent management and increase recruitment and retention.
Congress already gave the military services some new tools to do that. The services can now directly commission talented individuals and utilize more flexible promotion tracks.