WASHINGTON — The U.S. must act to preserve its edge over rival nations on artificial intelligence and cybersecurity, starting by embracing a new leadership position at the Pentagon, industry executives told Congress this week.
While the U.S. is ahead of China, Russia and other adversaries, deliberate and holistic efforts must be undertaken to maintain that lead, the officials with Alphabet Inc.’s Google and Microsoft Corp. said.
“We are ahead. We’re losing ground,” Andrew Moore, director of Google Cloud AI, said during a May 3 hearing of a Senate Armed Services cyber subcommittee. “I’m most worried about our structures. Bringing in massive-scale, superhuman automation means changing organizational structures and change management.”
Eric Horvitz, Microsoft’s chief scientific officer, offered similar guidance.
“The U.S. is leading in science, at the core principles and the creative applications, from my point of view,” he told the panel. “That said, these days technical advances spread around the world like lightning.”
The Department of Defense late last year created the chief digital and artificial intelligence officer position to accelerate all things digital at the Pentagon. The CDAO — soon to be Craig Martell, Lyft’s machine-learning executive — will oversee the Joint Artificial Intelligence Center, Defense Digital Service and the agency’s chief data officer, and will report directly to the deputy defense secretary. The office is expected to be fully operational this year.
Moore applauded the move Tuesday and wished Martell success.
“This is how we’re going to succeed, by having a centralized effort to put an artificial intelligence strategy across the whole DoD,” he said. “You cannot just magic AI on top of existing systems. You have to think about how you’re going to change operations.”
Moving forward, Horvitz said, there is a need to “get our hands dirty and work hard and then share ideas, insights, across the sectors.” Integrating scientific achievements into the Defense Department’s day-to-day routines will be crucial.
“We often think about AI … as on the battlefield, as kinetics,” Horvitz said. “But DoD is a huge operation, in peacetime and in war. The logistics, planning, predictive models, employment, back to health care, the VA system — all can benefit greatly by even basic applications of machine learning, predictions, diagnoses and planning.”
The Pentagon recognizes the value of AI and is actively pursuing and investing in it. Some, though, believe it’s moving too slowly or with incorrect priority.
“Make no mistake, our adversaries will capitalize on this technology, using AI to power attacks on our networks as well as increasing their ability to detect our intrusions on their networks and to respond quickly,” Sen. Joe Manchin, the West Virginia Democrat who heads the subcommittee, said at the hearing. Failing to prepare, he continued, “would be disastrous.”
A Defense Department assessment from 2018 warned that China and Russia are spending significant sums of money on AI for military applications. The investment threatens to erode U.S. technological and operational advantages, according to the document.
“Put simply,” Sen. Mike Rounds, R-S.D., said at the hearing, “our adversaries are going to use AI against us, so we must use AI to defend against them.”
Colin Demarest is a reporter at C4ISRNET, where he covers networks and IT. Colin previously covered the Department of Energy and its National Nuclear Security Administration — namely nuclear weapons development and Cold War cleanup — for a daily newspaper in South Carolina.