House lawmakers were warned Wednesday that artificial intelligence could soon be used by potential adversaries in military operations.
Jason Healey, a senior fellow on the Atlantic Council’s Cyber Statecraft Initiative, told members of a House panel with oversight of the Pentagon that he expects the capabilities to be developed in the next decade.
“There has been lot of speculation … about how soon it will be before robotic soldiers take the place of the fight in the kinetic world,” Rep. Mike Conaway (R-Texas) asked the panel of cyber experts. “How soon will A.I. supplant the need … for all these human beings to be able to defend these networks and do what we do?”
Healey answered that he expects the capability to be developed more quickly than anticipated.
Peter Singer, a strategist at the New America Foundation, said that artificial intelligence is among potential “disruptions” being developed in the realm of cyber conflict.
“We don’t want to fall behind," he said.
Healey expressed concerns about the possibility of artificial intelligence augmenting our adversaries’ offensive capabilities more significantly than the United States’ defense of its critical infrastructure.
“The part of it that particularly worries me the most is that on the defensive side many people are thinking that artificial intelligence, new heuristics, better analytics and automation are going to help the defense, that if only we can roll these things out faster we will be better and the system will be more stable,” Healey explained.
“I think that these technologies are going to aid the offense much more than it aids the defense because to defend against these kinds of attacks, you need your own super computer,” he continued.
Healey warned that while the Pentagon can afford computer systems necessary to defend against adversaries using artificial intelligence, small- or mid-sized enterprises that own U.S. critical infrastructure cannot.
“It leaves much of America undefended,” he said.
Rep. Mac Thornberry (R-Texas), who chairs the House Armed Services Committee, later suggested that while the U.S. might soon develop these capabilities, the government might not have sufficient policies in place to control how it uses them.
“It seems to me that we’re always a lot better at developing technologies than we are the policies on how to use them,” Thornberry said.
The hearing on 21st century cyber warfare focused on a wide range of issues relating to the U.S. military’s cybersecurity, including its vulnerabilities and capabilities in cyberspace as well as weaknesses in its cyber workforce. The panel also wrestled with what role the military has in using offensive and defensive cyber capabilities to protect critical infrastructure.
Lawmakers have paid more attention to potential vulnerabilities of government systems and critical infrastructure to cyber attacks in the wake of high-profile breaches, including the Russian government’s hack of the Democratic National Committee and the massive Office of Personnel Management breach tied to Chinese hackers.
Members of Congress have been pushing for a more comprehensive policy on how to deter and respond to cyberattacks and acts of cyber espionage and crime.