Lawmakers, DOD Officials Note Benefits of AI to Department’s Information Networks
During a hearing on securing the Defense Department’s information networks, senators and officials described how artificial intelligence and machine learning capabilities could strengthen the department’s cyber resilience.
Artificial intelligence could further safeguard the Department of Defense’s information networks—or DODIN—from cyber adversaries, lawmakers said during a Senate Armed Services Cybersecurity Subcommittee hearing on Wednesday, even as they expressed some concern about how nation states could potentially weaponize advances in AI to penetrate secure government networks.
Sen. Joe Manchin, D-W.V.—the subcommittee’s chairman—called the practical implementation of AI across DOD “a top priority” for both himself and Sen. Mike Rounds, R-S.D.—the panel’s ranking member—adding that “we’ve been speaking about that and learning a lot more about that.”
When asked by Manchin what tangible AI applications have been most successful for the department, John Sherman, DOD’s chief information officer, cited what the department has done “on preventative maintenance on helicopters, for example, using AI out at the tactical edge there to help our special operators on Blackhawk helicopter maintenance using AI.”
Rounds similarly expressed interest in learning more about the role “that ethical artificial intelligence will play in our cyber defenses” moving forward, particularly when it comes to countering aggressive cyber operations on the part of hostile nations that could be looking to weaponize technological advances for their own purposes.
“The AI is here already, and we're going to have to expand and we're going to have to take advantage of all of the opportunities, but defend against the challenges as well,” Rounds added.
Sen. Jacky Rosen, D-Nev., noted that adversarial nations and threat actors “could use AI in future cyberattacks in the United States,” including against DODIN, but that AI “also has great potential as a tool for the Department of Defense to hunt for malicious software and search for those irregular behaviors, if you will, that could indicate a presence of an intruder posing a threat to our DOD system.”
When asked by Rosen how DOD is leveraging advanced AI and machine learning models to improve DODIN, Sherman pointed to the department’s implementation of its zero trust framework as one approach to further incorporating these technologies into its networks, citing the framework’s pillars focused on automation and orchestration, as well visibility and analytics, “as we apply AI and ML to this.”
Sherman said DOD is committed to implementing the zero trust framework across the department by 2027, which he called “an ambitious yet critical milestone, given the geopolitical threats we face.”
“These modern threats demand that we maintain a relentless focus on eliminating technical debt,” Sherman added. “All of our systems, be they for weapons enterprise IT, command and control, business systems or defense critical infrastructure must be equipped with the most modern cyber defenses that can stand up to savvy and determined state and non-state actors.”
Lt. Gen. Robert Skinner—the director of the Defense Information Systems Agency and the commander of the Joint Force Headquarters responsible for operating and securing DODIN—noted that DOD “is using AI in multiple points within the DODIN,” including leveraging AI models models “to find something that we may have missed, initially, to holistically get after the cybersecurity threats.”
“At our different endpoints, many of the products today have artificial intelligence already embedded in them,” Skinner added. “So even as we're purchasing them, we're leveraging and they're at our boundary. We're actually leveraging artificial intelligence to find those irregularities.”
The panel’s members expressed an interest in learning more about DOD’s use of, and defense against, emerging AI capabilities moving forward, and Manchin said that the subcommittee’s next hearing—which has not yet been scheduled–will be “focusing solely on AI.”