Can the EU and its allies cooperate with China on AI safety standards?

With geopolitical tensions between the West and China at an all-time high, the explosion of interest in AI could not have come at a more awkward time, given the global effort needed to work together on setting safety standards.

As governments throughout the world race to regulate the nascent technology, the focus on international cooperation, particularly with China, is intensifying.

It was a key topic at Euronews’ International AI summit in Brussels this week, with experts and analysts attending the event arguing it could prove difficult to cooperate with Beijing on AI standards.

Anu Bradford, a professor of law & international organisation at Columbia Law School in New York, agreed with this assessment.

“I think it’s fair to say that the geopolitical reality is not very conducive to deep, meaningful cooperation,” she told Euronews.

“There is very much an escalating tech war between the US and China. There’s a tremendous race for technological supremacy. There’s concern who is the economic power, the technological power, the geopolitical power. 

“And there are also vast ideological differences,” Bradford added.

“So, it is hard to see the US and the EU agreeing with China on particularly meaningful, substantive rules around AI.”

Cooperation still possible

But there is some hope amid all the geopolitical tension. Last week saw the historic Bletchley Declaration in the UK where the US and China both signed a Declaration on AI Safety.

According to Rebecca Arcesati, a lead analyst at the Mercator Institute for China Studies – which has been sanctioned by Beijing – there is also optimism in the business-to-business area.

“I’m very optimistic when it comes to what can be done between companies, between engineers, Chinese companies that are already playing a really active role,” she said in an interview.

“As long as these conversations can continue and even maybe be facilitated by governments, despite the geopolitical tensions, I think there will be progress going forward.”

Arcesati did add, however, that at a governmental level, cooperation will likely remain elusive for now.

“At the level of government-to-government engagement, we still have to come to terms with very different political systems, and that makes engagement very difficult with China,” she said.

“China in international bodies, such as the United Nations, has a specific agenda where it would like to legitimise its own domestic approach to AI governance and that is an approach where, for example, individual human rights are disregarded very much whenever state security needs to be protected.

“That obviously poses a dilemma for liberal democracies, who would like to engage with China.”

West must first come together

Talk of cooperation with China could be premature though.

Dragoș Tudorache, a Romanian MEP who is in charge of the EU’s AI Act within the European Parliament, told Euronews that AI cooperation with China should, for the moment, be secondary.

First, the collective West needs to get its own house in order.

“When addressing governance at a global stage, when addressing standards at a global stage, there is an inevitable dialogue with China,” he said.

“But what I’ve always said is that first and foremost we have to make sure we are first in line – us the democracies that understand technology in the same way and its role in society – we have to first get ourselves as convergent as we can, as aligned as we can.

“And then have a proper conversation with China to make sure that we can address as many of the other, let’s say, bigger risks, including geopolitical risks, in a framework that also includes them.”

Next week could see another historic moment in the West’s cooperation with Beijing.

US President Joe Biden will likely meet with Chinese leader, Xi Jinping, which could lay the groundwork for future cooperation on artificial intelligence.