The near-term threat to the US from China exploiting AI tech isn’t from Terminator-style cyborg assassins popularized by Hollywood, analysts say. Rather, the immediate danger comes from more mundane uses of AI — such as analysis of personal information and “open-source” data in the US for espionage and other intelligence activities.
“The more likely, impactful risk of AI will involve something like sifting through publicly available data to better target people,” John Schaus, an Asian security expert and former Pentagon official told Asia Times.
Schaus, a fellow in the international security program at the Center for Strategic and International Studies, notes that when such personal data is gathered by credit card companies or Facebook, it’s categorized as “marketing.” But when done by a foreign state or government-controlled entity, it amounts to intelligence collection.
He says players like China and Russia can tap AI to engage in ID theft, invade data banks and bank accounts and create other ways to support intelligence operations.
“The more likely, impactful risk of AI will involve something like sifting through publicly available data to better target people”
Such concerns emerged in June 2015 when the US Office of Personnel Management was the target of a data breach that compromised the records of four million federal employees. The unidentified hackers, who were allegedly traced to China, stole Social Security numbers and may have taken detailed security-clearance-related background information.
On the military side, both the US and China are keen on using AI and machine learning to process video and other data for drone strikes.
The US already incorporates the technology for drone missions against the Islamic State. China also wants to tap AI for domestic surveillance and missile systems.
Schaus notes the object and facial recognition feature of AI technology can be directed at civilian areas to detect patterns in daily life by tracking car, foot or air traffic, or collecting signals intelligence from given areas.
Nations fear AI arms race
But Hollywood’s vision of robotic tanks, planes and soldiers is still in its early stages. Schaus says nations like the US and China are wary of unleashing an AI “arms race” that opens the gates for AI-driven weapons and robotics. Top concerns include the inherent risks of using non-human decision makers in advanced weapons systems, as well as cost and retooling issues.
“No nation wants an AI arms race,” Schaus said.
But Winn Schwartau, a noted cyber warfare analyst, says the race is inevitable. “Every technology will be weaponized,” said Schwartau, who coined the phrase “Electronic Pearl Harbor” in the 1990s.
Doug Tsuruoka is Editor-at-Large at Asia Times