‘Deepfake geography’ could fool military on battlefield with lethal consequences
Artificial intelligence experts have already warned of the dangers that deepfakes could pose a serious threat to democracy.
But a new use for the image-faking technology has been exposed by researchers.
“Deepfake geography” could be used to create misleading satellite images for propaganda purposes or even potentially to get a tactical advantage over an enemy on the battlefield.
Bo Zhao, assistant professor of geography at the University of Washington and lead author of the study, says in a new study published on April 21 that “deepfake geography” could grow to become a significant problem.
“This isn’t just Photoshopping things. It’s making data look uncannily realistic,” he writes in the journal Cartography and Geographic Information Science.
“The techniques are already there," he says "We’re just trying to expose the possibility of using the same techniques, and of the need to develop a coping strategy for it.”
As Zhao explains in his research paper, mapmakers have been inserting small deliberate errors into their work for centuries to prevent copyright theft. But people’s increasing reliance on satellite image means that the deepfakes will affect more than just rival geographers.
For example, the detention camps in which China is holding over 1.5 million Uyghurs, Kazakhs and other minorities were first identified via satellite images. A well-placed deepfake image could literally wipe these camps off the map.
He demonstrated the technique by using satellite images of Tacoma in Washington State overlaid seamlessly with streets from Chinese capital Beijing.
Todd Myers, automation lead for the CIO-Technology Directorate at America’s National Geospatial-Intelligence Agency says that this technology is already under development in China: “The Chinese are well ahead of us. This is not classified info,” he said in March 2019.
“The Chinese have already designed; they’re already doing it right now, using GANs—which are generative adversarial networks—to manipulate scenes and pixels to create things for nefarious reasons.”
Myers said that, for example, editing satellite photos to “move” a bridge could easily wrong-foot an enemy.
“From a tactical perspective or mission planning,” he explained, “you train your forces to go a certain route, toward a bridge, but it’s not there. Then there’s a big surprise waiting for you”.
Even battlefield drones sent out to reconnoitre an area before troops move in could be fooled by AI-created camouflage.
DARPA’s Hava Siegelmann demonstrated how simple stickers can be added to road signs. They are barely noticeable to human drivers but can completely confuse machine vision systems.
Andrew Hallman, head of the CIA’s Digital Directorate, said the entire concept of truth is under threat. “We are in an existential battle for truth in the digital domain,” he said.
“That’s, again, where the help of the private sector is important and these data providers. Because that’s frankly the digital conflict we’re in, in that battle space…This is one of my highest priorities.”
Source: Read Full Article