It’s no secret the intelligence community is pursuing applications of artificial intelligence to further its national security mission. But one of the underlying challenges for the IC is the scattered and often-unorganized patchwork of data across the 18 intelligence components.
Last year, Director of National Intelligence Avril Haines signed out a two-year strategy aimed at overhauling the IC’s data management practices. And last week, the office of the director of national intelligence finalized a new “data reference architecture” that aims to enable a “distributed data ecosystem,” IC Chief Data Officer Lori Wade said at the Defense Department Intelligence Information System conference in Omaha, NE, on Tuesday.
“We need to get to a place where we have domain ownership and decentralized stewardship,” Wade said. “This data ecosystem needs to pull in the data services that we have. It needs to pull in all of the capability that we have.”
The goal isn’t to centralize all the IC’s data under one system, Wade added.
“We can’t afford to do that,” she said. “We can’t afford to keep duplicating data. We need to leave it where it is, and we need to go with the data mesh principles. That’s a data management approach of where we think about leaving data where it is, and combining the data at the semantic layer . . . We think of data as a product.”
The reference architecture is unclassified, Wade said, and will be shared across the IC, DoD, private industry, and the “Five Eyes” intelligence partners.
“This is the call to everyone in this room: Now we need to implement it. We need to work together,” Wade added.
The push to improve data management comes as the intelligence community prioritizes the adoption of “AI at scale” under a new IT roadmap released this year. The roadmap also shows that between fiscal 2026 and 2029, officials plan to establish “AI enabling services at scale,” including a model repository and training data.
“Secure, generative, and predictive AI can reduce the time for intelligence insights from days or weeks to mere seconds,” the IT roadmap states.
Meanwhile, President Joe Biden last week issued a sweeping national security memorandum aimed at accelerating the adoption of AI while addressing security and safety risks.
In addition to finalizing the new data reference architecture, Haines in June also signed out a new intelligence community data management directive. The directive “establishes policy for the standardization and governance of data that the Intelligence Community (IC) collects, acquires, creates, holds, stores, accesses, or uses to maximize and provide consistency in its usability and interoperability.”
To help implement its revised data practices, ODNI earlier this year established a “data operations group” to test out “real mission challenges,” Wade said.
“How are we going to do a mission sprint to stress test this?” she said. “How do we bake this in to how we’re doing that end to end data management?”
ODNI will also host an AI and Data Summit in January to further wring out some of the challenges with government and industry partners.
“We can’t wait and just think we’re going to jump out one day and be like, ‘Oh, AI saved us,’” Wade said. “It’s not going to happen unless we plan for it, unless we implement this planning that we’ve put together. A lot of smart people in this room have participated in that, and now we need to move out, and we need to invest in all of those areas . . . We got to work together. We have to innovate, but we can’t do it at the expense of all of the foundational work that has to be in place, and we need the workforce to do it.”
Copyright
© 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.
