US raises concerns with China over AI misuse

The White House said on Wednesday that in their first formal talks on the topic, U.S. and Chinese officials talked about their worries about China‘s “misuse” of artificial intelligence. This comes as the two superpowers try to avoid a fight over the rapidly developing technology.

While meeting with their Chinese counterparts in Geneva on Tuesday, the U.S. delegation also stressed the need to “maintain open lines of communication on AI risk and safety as an important part of responsibly managing competition,” according to the White House.

“The United States also raised concerns about the misuse of AI, including by the People’s Republic of China,” Adrienne Watson, a spokesperson for the White House National Security Council, said.

U.S. officials are worried about China’s access to AI technology because they think Beijing could use it to hack other countries’ elections, make bioweapons, and attack them online.

The U.S. has already told China and Russia that only people, not AI, would decide when to use nuclear weapons. The State Department has already instructed them to do the same. President Joe Biden‘s administration is about to start a new effort to protect U.S. AI from China and Russia. Sources say that the administration is planning to put fences around the most advanced AI models, which are the core software of AI systems like ChatGPT.

Last week, a group of lawmakers from both parties in the United States introduced a bill that would make it easier for the Biden administration to control the export of AI models. On Wednesday, a group of senators from both parties in the U.S., led by Majority Leader Chuck Schumer, called for a big increase in government funding for AI research while they talked about new legal protections.

American AI chips and the tools used to make them aren’t going to China as much as they used to. The Biden administration also aimed to mandate that U.S. cloud companies notify the government when customers from other countries utilize their services to train AI models capable of computer attacks.

China relies on a lot of open source models created in the West, like the Meta Platforms and “Llama” series. Also, a lot of high-tech American AI chips are making their way to China.