Going forward, the Pentagon will focus on agile artificial intelligence adoption throughout the department to keep up with the ever-evolving technology.
On Thursday, Deputy Secretary of Defense Kathleen Hicks unveiled DoD’s new AI strategy at a press event. The strategy follows President Joe Biden’s executive order on AI earlier this week, which ordered the DoD to create a pilot program to explore how it can use AI to protect the nation’s national security systems and networks.
“Today, we’re releasing a new data, analytics and AI adoption strategy, which not only builds on DoD’s prior year AI and data strategies, but also includes updates to account for recent industry advances in federated environments, decentralized data management, generative AI and more,” Hicks said.
She said the strategy is meant to create a foundation for the adoption and usage of data, analytics and AI across the Defense Department, and promote speed, delivery, learning and responsible development.
“Our task in DoD is to adopt these innovations wherever they can add the most military value. That’s why we’ve been rapidly iterating and investing over the past two plus years to develop a more modernized, data-driven and AI-empowered military now,” Hicks said. “As we focused on integrating AI into our operations responsibly and at speed, our main reason for doing so has been straightforward, because it improves our decision advantage.”
However, the Defense Department cannot use many commercially available AI tools because they currently do not align with the department’s ethics.
“As commercial tech companies and others continue to push forward the frontiers of AI, we’re making sure we stay at the cutting edge with foresight, responsibility and a deep understanding of the broader implications for our nation,” Hicks said. “For instance, mindful of the potential risks and benefits offered by large language models and other generative AI tools, we stood up Task Force Lima to ensure DoD responsibly adopts, implements and secures these technologies. Candidly, most commercially available systems enabled by large language models aren’t yet technically mature enough to comply with our ethical AI principles, which is required for responsible operational use.”
The DoD has found areas that would make be useful for generative AI.
“We have found over 180 instances where such generative AI tools could add value for us with oversight, like helping to debug and develop software faster, speeding analysis of battle damage assessments and verifiably summarizing tax from both open source and classified data sets,” Hicks said.
The AI strategy will help the DoD unify, synchronize and scale AI across the enterprise. It focuses on creating an environment for DoD leaders and personnel to effectively use AI, data and analytics. A key component is the agile adoption of data, analytics and AI across the department to help leaders make better, faster decisions.
“This is not a capability development strategy,” Craig Martell, the department’s chief digital and AI officer, said at a press conference. “Technologies evolve, things are going to change next week, next year, next decade and what wins today, might not win tomorrow. Rather than identify a handful of AI-enabled warfighting capabilities that will beat our adversaries, our strategy outlines the approach to strengthening the organizational environment with within which people can continuously deploy data, analytics and AI capabilities for enduring decision advantage.”
There are several goals outlined in the strategy, such as improved data sets and infrastructure, more partnerships with outside groups and removing internal barriers to help the department keep pace with adopting advancing technologies.
Specifically, the strategy focuses on DoD’s AI hierarchy of needs, with good data as a foundation to responsible AI usage. This is followed by analytics and responsible AI. The hierarchy is intended to be a logical starting point, not a temporal one.
“Logically speaking, if you do not have high quality data, analytics is a fool’s errand,” Martell said. “If you do not have high quality data, AI is a fool’s errand.”
A key element of the strategy is data shareability to help the Combined Joint All-Domain Command and Control, or CJADC2. Working with industry will be key to remove data silos.
“Now that we have clarity about what we want to build, we can’t build that. It’s not something that government should build,” Martell said. “We have to work very closely with our industrial partners. And what we have to figure out now is how do we get our industrial partners to work with us in a way where they help us build out this open standard data layer, and the data that they provide isn’t locked up in a silo. That’s going to be our biggest challenge. If we end up having providers continually locking data up in silos and not in this data mesh that allows for free discovery and accessibility of the data, then that’s going to be a blocker, so that has to be a real challenge we have to break through.”
Martell added that conversations with industry were promising about this new thought process.
As part of the new strategy, DoD will have a decentralized network across its data providers and users instead of a single office handling a centralized network. The different services can choose their own provider that best suits their needs as long as they are adhering to the department’s established best practices. This should enable data sharing across the department to further CJADC2.
An implementation strategy will follow in the coming months, which will be a set of best practices.
“Each of the services have wildly different needs and they’re at they’re wildly different points in their journey and they have wildly different infrastructure,” Martell said. “So we’re going to insist on patterns of shareability, patterns of accessibility, patterns of discoverability. And how those are implemented, we’re going to have a lot of variants for.”
DoD last released an AI strategy in 2019 and updated it in 2020. The older strategy established the now-defunct Joint AI Center to fulfill this vision. However, the CDAO subsumed JAIC when it was established last year.
“In 2018, the then JAIC, focused on building a centralized AI-ML pipeline and that makes a lot of sense for 2018 because even industry hadn’t yet figured out how to deliver that as a product to customers,” Martell said. “But in 2022 every one of the major vendors delivers a robust and industrial scale machine learning operations pipeline, so there’s really no need for us to build that internally.”
The 2023 strategy was developed by the CDAO and unifies the prior strategies and guidance. It will also take other DoD tools and resources like Task Force Lima — which looks into responsibly using generative AI — into consideration as the department moves forward with its use of AI.
Kirsten Errick covers the Defense Department for Federal News Network. She previously reported on federal technology for Nextgov on topics ranging from space to the federal tech workforce. She has a Master’s in Journalism from Georgetown University and a B.A. in Communication from Villanova University.