Agencies are drowning in cyber domain data, and when it comes to enabling rapid response, some of their biggest challenges include quality and accessibility of data, Federal defense and intelligence officials said this week.
At ManTech’s TechShare 2024 conference in Washington on May 1, officials from the Department of the Navy and National Security Agency (NSA) said they are leveraging AI to zero in on operationalizing actionable, real-time data that enables them to respond quickly to advance the mission.
“Looking at AI, we know we want to be a data-driven organization, so that means pulling in more data to do those analytics, but it’s managing those data where I think AI can really help us to comb through that large amount of information that continues to grow,” said Alvin “Tony” Plater, chief information security officer for the U.S. Navy. “How can we take AI to, in fact, get after the information we need to help us drive the decisions and give us the advantage?”
“I really see AI as playing that role where we are actually able to make that data decision and to do it at speed and scale,” he added.
Colin Crosby, who is service data officer at the U.S. Marine Corps and deputy chief data officer for the Navy, said AI has been the number one goal for the service across the spectrum. According to the data expert, AI has helped marines reduce planning from days down to minutes.
“We have AI at the enterprise level. We’re integrating AI in our records management program. We’re integrating AI in our information and knowledge management program,” Crosby said. “We’re enabling AI in our enterprise data management. We’re using AI for sharing our network. We’re enabling AI for target identification, target classification, and with the appropriate governance, at some point target prosecution.”
Because the NSA is dealing with broad amounts of data, the Deputy Chief for the agency’s Cybersecurity Collaboration Center, Atiya Yearwood, said AI tools are helping with that volume challenge.
“AI has been helping NSA and will continue to help NSA reduce down the amount of information just to focus our human resources on where we actually need an analyst brain to get information to a policymaker,” she said, adding, “just the number of documents that we may be sifting through.”
“From a cybersecurity perspective, our AI, we’re looking at ways for it to help us with malware detection when we’re looking at terabytes of data, and sometimes we just need this one little piece to help us quarantine information, set it aside, and protect the system,” Yearwood said.
All three officials agreed that the future of AI will involve both enabling and impeding cybersecurity.
“From the helping perspective, just more to that point of when we’re looking at malware detection, we’re able to use AI in that way. But the flip side of that is that AI can be used to enhance or also generate malware,” NSA’s Yearwood said. “We’re seeing that AI can be used to subvert multifactor authentication, or it can also just subvert any malware detection mechanisms that we have in place.”
“We’re also very concerned about a space where our AI is fighting other AI just to keep outpacing each other,” she said. “And a large concern that we have is just with the proliferation of AI comes a human dependency on it. And also with that dependency could potentially come resting on our laurels. And so that’s a big concern for us. And so, we’ve been very focused in the center on AI security.”