Warnings about spying, hacking threat from Chinese AI

In a new Australian Strategic Policy Institute report, Simeon Gilding says AI products used or created by Chinese companies can create security risks and expose Australians to spyware or hacks.

The institute senior fellow says the concern isn't China's AI capabilities, rather companies that need to succumb to the authoritarian government's wide-ranging security laws.

"They're subject to direction from PRC security and intelligence agencies," he writes.

The risks include opening Australians up to espionage through surveillance and data theft, as well as the sabotaging of systems and services.

The use of AI to conduct foreign interference, including the mass spreading of mis and disinformation, also poses a threat.

"So we in the democracies need to ask ourselves against the background of growing strategic competition with China, how much risk are we willing to bear?" the report says.

Mr Gilding posed three possible solutions to reduce the risk - auditing current systems, bringing in experts to find holes in security systems by using "a thief to catch a thief" and regulating the emerging technology.

Regulation could include prohibiting Chinese AI technology on certain systems, such as national security networks or government devices, or banning it entirely.

But he notes a blanket ban would be costly and disruptive.

"Many businesses and researchers in the democracies want to continue collaborating on Chinese AI-enabled products because it helps them to innovate, build better products, offer cheaper services and publish scientific breakthroughs," he said.

Mr Gilding said regulation was needed to prevent further security threats.

"It is a balanced measure in a world in which China is neither at peace nor at war with us," he said.

"We should be vigilant about the balloons in the sky, but we should think harder about the ghosts in the machine."

Australian Associated Press