Trump-Xi meeting brings AI arms race into focus

Since last week, researchers from several of the companies leading the development of artificial intelligence have come forward to warn that the technology could pose an existential threat to humanity. These researchers argue that the quest to develop more and more powerful AI models is creating an attitude of indifference about safety and pushing aside any concerns that AI could quickly escape human control. Industry leaders have criticized Trump’s push for global AI supremacy and say he has prioritized the desire to develop AI quickly rather than developing it responsibly.

These recent warnings from industry insiders have added to a growing public skepticism over AI. There is a growing perception among the public that tech oligarchs are becoming indifferent to the fate of humanity and cannot be trusted to develop AI responsibly. At the same time, bipartisan resistance is growing against the rush to build AI datacenters all over the country, despite the drawbacks and risks to local communities.

President Trump himself has repeatedly scorned both the public and industry voices calling for the government to rein in runaway AI development. Trump recently called concerns about AI’s dangers to humanity “a hoax” and said that communities who don’t want AI data centers want to remain “backwards and poor”.

The new Cold War

Next week, Chinese President Xi Jingping will meet with Trump in Washington, DC. When the summit was scheduled in May, the two leaders had agreed that talks about AI safety would be on the agenda. However, given Trump’s recent rhetoric, it may now prove more difficult for the two to find common ground on this topic.

In recent statements, Trump has repeatedly cited the need to “win” the AI race against China. Westerners who support continued, rapid development of AI have pushed the debate into the geopolitical arena, claiming an imperative to maintain a technological, military and economic edge over China. Chinese officials, on the other hand, have described US export controls on chips as an attempt to thwart the development of China’s AI sector. Both sides frame the race to develop AI in existential terms which echo the rationale for maintaining the nuclear arms race during the Cold War. The result is a mutual distrust that makes any cooperation or mutual understanding on AI safety near impossible.

Chinese officials have recently voiced security concerns about AI, specifically the possibility that it could be used to hack state systems. Fear of an extinction-level threat from AI are less prevalent in China’s AI sector, in part because China’s AI models are not as advanced as the US. Western AI companies have reported numerous breaches where their AI systems have carried out unauthorized hacks of other systems. If such events have occurred in China, they are not public knowledge.

Nevertheless, the Chinese do take the threat of AI escaping human control seriously and are already shaping policy to mitigate the risk. At a recent conference, President Xi warned of the rapid advance of AI technology and said that it must always remain within human control.