The Defense Information Systems Agency (DISA) is looking to “automate and elevate” its data in order to better adapt to a dynamic cyber environment, according to Darrell Fountain, chief of the Cyber Security Services Provider (CSSP) Program at DISA.
At the WEST 2024 conference on Feb. 14 in San Diego, Calif. – co-sponsored by AFCEA International and the U.S. Naval Institute – Fountain explained that DISA is leveraging artificial intelligence (AI) and machine learning (ML) to help it sift through mountains of data.
“We have to automate and elevate our game,” Fountain said. “We have to automate the mundane, routine tasks that our analysts are burdened with today – chasing down [Indicators of Compromise], looking for artifacts, and trying to sift through those. I need to automate so I can move that analyst up the value chain. I need to move him up where automation isn’t quite there yet.”
DISA’s CSSP Program has a pretty large partner base, as it supports over 450 mission partners across the department. To support its many partners, DISA utilizes several different cloud providers “who all generate data in different ways,” Fountain said.
This poses a big challenge to DISA, he explained, as it can be difficult to analyze multiple relevant security data pools to find “good data.”
“We need to set the conditions for how we handle our data because data is the key,” Fountain said. “Everything tastes better when you sprinkle AI/ML on top of it, right? … But what is the engine of AI/ML? Good data. You’ve got to have good data. What is the agency struggling with? Good data.”
“I have to be consistent, persistent, and disciplined in how I manage that data,” he continued. “So, I set the conditions to use AI/ML, and when I get some of that AI and I sprinkle it on top of my analysts like pixie dust, I need that to work with me. I don’t want AI/ML that I work for – I need AI/ML that works for me.”
Fountain stressed that automation will help analysts to “regain control” of relevant security data. While he may not know what he’s looking for in the data today, Fountain said it’s crucial to “harvest data in creative ways” so that AI and ML will have the opportunity to do the work in the future.
“I need to create the opportunity for AI to be my partner,” he said. “If I don’t gather data in a persistent, consistent, disciplined way, then AI really can’t help me – it can guess, but it can’t be the keys to the kingdom that I think it will be.”
Looking towards the future, Fountain said he predicts that the era of generating your data with organic tools will go away as AI and ML take over. “Put all of your data in a big pile, and then let the wonders of modern automation sift through it in ways that the gray matter between my ears cannot and call out of it what we need,” he said.
