The Pentagon's Chief Digital and Artificial Intelligence Office (CDAO) will focus on enabling AI tools for combatant commands next year, according to CDAO Chief Technology Officer William Streilein at ATARC's Federal AI & Data Summit yesterday.
While the military services are already adopting AI capabilities, the combatant commands need help leveraging the services' capabilities to enable the joint force.
While there are teams of experts available within the Defense Department to help combatant commands deploy AI, the CDAO wants the commands to be able to organically identify, develop and deploy their own capabilities to help transform how they pursue their missions.
"I don't want to say [the combatant commands are] left behind, but [they] are basically in need of transforming, and so what we'll be able to talk about next year is some strides there," Streilein said Thursday.
The CDAO was formed out of four previously-existing organizations responsible for AI and data within DOD: the JAIC (Joint Artificial Intelligence Center), the Chief Data Office, the Defense Digital Service (DDS), and ADVANA, a data analytics platform. The goal of joining these four organizations together was to streamline DOD's efforts to accelerate transformation related to data, AI and analytics. The CDAO reached full operating capability in June 2022 and will spearhead all AI and data-related initiatives moving forward.
"The goal is not to create one monolithic system for all of DOD but to federate, to enable systems that already exist to share data and work across each other," Streilein said.
To be AI-ready by 2025 and AI-competitive by 2027, and before the department can leverage analytics and AI, leadership needs to focus on quality data, according to Streilein.
"We have to get our data in order because the data will supply you the fuel, if you will, for the analytics," Streilein said.
"Teaching best practices related to data - that is probably the most important thing… Data gets created all the time. And it's just so easy to treat it and not put the right labels on it," Streilein said.
To achieve that goal and become data-centric, all the data needs to get to a place where it is VAULTIS, or visible, accessible, understandable, linked, trustworthy, interoperable and secure.
"If you can make your data VAULTIS then you are hopefully AI-ready. That's great if you can say your new data's VAULTIS, then you're not creating new bad, and you're moving things forward, and so incrementally we're getting there," Streilein said.