To Defeat Enemy Drone Swarms, Troops May Have to Take a Back Seat to Machines, General Says

The Army's top modernization official said Monday that the Pentagon may need to relax its rules on human control over artificial intelligent combat systems in order to defeat swarms of enemy drones, which are often moving too fast for soldiers to chase.
All branches of the U.S. military have expressed an interest in using artificial intelligence (AI) for faster target detection. However, the Department of Defense has so far stressed that people, not machines, will always make the decision to fire deadly weapons.
As small unmanned aerial systems (UAS) proliferate around the world, army modernization officials are realizing that swarms of fast-moving drones are difficult to defeat without advanced technology.
Read on: Futuristic 'Defiant X' in action to become the army's future long-range assault helicopter
"It just gets very difficult talking about swarms of small drones - not impossible, but more difficult," Army Futures Command chief General John Murray told an audience Monday during a webinar at the Center for Strategic & International Studies.
Murray said Pentagon leaders may need to hold talks about how much human control over AI is required to be safe but still effective against threats like swarms of drones.
"If you're defending against a swarm of drones, a human may have to make that first decision, but I'm not sure a human can keep up," he said. "How much human engagement do you actually need if you don't make fatal decisions from a human perspective?"
The army is experimenting with AI for faster and more accurate target detection. In the past, the service's mechanized combat units tested potential new tank gunners using index cards with images of armored vehicles in service around the world.
"New gunners got tests with index cards - pictures of different armored vehicles; they went through 15, 20, 25, 30 of them. If a soldier got 80 percent of it right, he was put in the gunners' seat of a very deadly vehicle," said Murray.
During an army exercise called Project Convergence at Arizona’s Yuma Proving Ground in September, modernization officials tested AI-enhanced systems in combination with near-earth satellites and other technologies to dramatically reduce the time it takes to identify, track and destroy incoming airborne threats.
"The operators we trained at Project Convergence were routinely 99 to 98 percent correct, so AI has the ability in many ways to make us safer," said Murray. "When you think of a future battlefield, in the past I've only described it as being hyperactive. I think decisions have to be made at such a pace that it becomes incredibly difficult for a human decision maker to keep up with them."
- Matthew Cox can be reached at matthew.cox@military.com.
Related Topics: Congress wants to step up army's efforts to develop counter-drone weapons

Last News

Stock Slide Continues on Growing Bond Concerns

Prince Harry tells all to James Corden on Archie's first word, his views on The Crown and family life with Meghan

Native American nominee's grilling raises questions on bias

A triple murder suspect told police he killed his neighbor and tried to feed his family her heart 'to release the demons'

Biden USTR nominee Tai vows to end trade 'race to the bottom,' counter China

Ashlee Simpson Ross' Husband Evan Ross Recalls Daughter's Fury About Getting a Baby Brother