Home » Despite risk, AI research is ‘entangled’ with China – Study

Despite risk, AI research is ‘entangled’ with China – Study

EUROPE-CHINA

European Union Commission President Ursula von der Leyen and other European leaders have referred to ‘derisking’ relations with China, including research collaboration in the key field of artificial intelligence, but Europe is still ‘entangled with China’ when it comes to AI research, according to a new report.

AI research collaboration between China and Europe has grown significantly in the five years from 2017 to 2022, despite temporary disruption from the COVID-19 pandemic, according to the report from the Mercator Institute for China Studies (MERICS) in Berlin.

“AI research collaboration is valuable for China and for Europe. China’s growing importance in the field makes it a critical partner,” it states.

Furthering military and authoritarian aims

Yet some of these collaborations are clearly benefitting China’s military and the Chinese government’s authoritarian aims. And there are also concerns that the deployment of ever more capable AI systems, like generative AI GPT-4, could “pose serious social and even existential risks”, unless effort is devoted to safety, alignment and other risk research, states the MERICS report.

The MERICS study, based on analysis of co-authored papers in the OpenAlex open source code repository, found that over the five-year period, some 16,386 papers were co-authored with peers affiliated with Chinese entities either directly controlled by the military or having close ties to it – roughly half of all papers in the five-year sample.

At the same time, the use of AI by the Chinese Communist Party to strengthen its authoritarian rule in regions such as Xinjiang or Tibet, overshadows cooperation, the report warns.

Around 87 papers were published with China’s military, the People’s Liberation Army (PLA), itself, according to the report.

The ‘dense and diverse entanglements’ – which often go back to initiatives by individual scientists, rather than institutional partnerships – that raise ethical or national security concerns “coexist with benign outputs and applications”, it says.

“The same AI model can be used for good and bad purposes, making drawing red lines difficult,” states the MERICS report.

Dangers of decoupling

But with the growing ‘entanglement’ in AI research between Europe and China, simply decoupling could harm Europe’s own AI research endeavour, the report argues.

Rather, “a recalibration of collaboration with China is needed, based on an informed assessment and management of risks, as opposed to a severing of ties that would compromise the benefits from exchange,” states the report.

“A broad severing of AI ties is not realistic, therefore European governments have an important role to play in enabling and preserving safe, ethical, and beneficial collaboration,” states the report.

“Ultimately, European governments will have to accept it will not be possible to prevent any leakage or ‘spin-out’ from commercial to military or surveillance AI, because that would mean halting collaboration across the board. Even then, the results of most European research would be published and still accessible [to China],” said Rebecca Arcesati, one of the authors of the report.

The report points out that this should temper expectations that “blunt instruments such as export controls” can stop the dissemination of knowledge. “However, meaningful knowledge transfers may also occur during the research process itself,” it adds.

State of China-Europe collaborations

Arcesati pointed to other studies which suggest that geopolitical tensions affected research ties between the US and China between 2020 and 2022.

“But it’s not the case for China- Europe,” she told University World News. “Over the same time, they [joint research papers on AI] continue to increase for all EU member states, with the exception of Bulgaria and Lithuania. And they also increased for [non-EU members] Norway, Switzerland and the United Kingdom.

“Geopolitical tensions and the COVID pandemic did not meaningfully alter the pace of Europe-China collaboration. Obviously, for the year 2020 for AI co-publications, co-authorship increased more slowly. But after 2020, you see a rapid rise again,” she added.

Despite the alarm being raised repeatedly by the US administration and by intelligence agencies in several European countries, China and Europe were becoming more important AI research partners between 2017 and 2022, with many joint research outputs having military or surveillance applications.

“These include, among others, collaborations on target tracking, cybersecurity and biometric recognition with risky entities in China,” the report notes, warning that European researchers should “consider the prospective partners, the specific research, and its potential end-uses”.

UK: China’s second largest AI partner

After the US, the UK is the second largest country partner for China in AI research. Arcesati noted that almost half of all Europe-China collaborative papers on AI were with the UK. In 2021, 21% of the UK’s entire AI research output comprised papers co-authored with Chinese partners, “which may indicate an uneven dependency in terms of talent and knowledge production”, according to the MERICS report.

After the UK comes the EU collectively. European-Chinese ties in commercial AI innovation are also significant. Several European companies have AI labs in China, while Chinese firms offer products and services in Europe.

Corporate R&D partnerships aim to use AI for a variety of applications, like drug discovery, and China is a market for some European AI products and services.

Since research and applications are often closely intertwined, such access matters not only for commercialisation but also for R&D advances, according to the MERICS report. Research collaboration has implications for Europe’s economic competitiveness and national security.

“Policymakers, industry, the research community and civil society worry about Europe becoming a second-tier player whose innovations and talent feed China and the United States. The EU and several member states have committed billions of euros in investment to catch up,” states the report.

“Whether technology, knowledge and capital transfers from Europe are helping China outcompete advanced economies in AI, R&D is the object of debate. As China becomes a top player in AI research, however, collaboration has become more of a two-way street. European countries have an interest in maintaining connections to be part of cutting-edge research efforts happening in China,” the report says.

Uneven partners

In terms of the quantity of research papers published, China and European countries are becoming more important partners, albeit at a different speed, indicating an uneven partnership.

The share of AI papers co-authored with China grew from around 5% of Europe’s total in 2017 to around 10% in 2021, whereas it went from 6% to 8% of China’s total. “China-authored papers are increasing at a higher rate than co-authored papers, while Europe’s output grew slower than both China-authored and co-authored papers,” the report says.

Among the papers whose funding source was disclosed (roughly half of the papers in the analysis), 80% were China-funded. The overall disclosed funding came from Chinese government programmes in 60% of cases, and most of the rest from universities.

Some Chinese funding was more obviously oriented to strategic aims. For example, seven papers had funding from China’s National Defence Basic Scientific Research Programme.

UK-China

The United Kingdom was involved in producing roughly one-sixth of Europe’s AI papers but almost half of European-Chinese co-authored papers. Between 2017 and 2021, the share of UK papers co-authored with Chinese partners rose from 11 to 21%.

By contrast, co-authorship with UK partners accounted for an average 3% of China’s English-language AI research output over the same period. “This indicates that authors with Chinese affiliations contribute more to the United Kingdom’s AI research output than the other way around,” states the report.

“The imbalance may reveal an emerging dependency, suggesting that limiting collaboration and losing Chinese talent would come at a cost,” it adds.

British institutions with a Chinese campus or cooperation with a Chinese university have the largest number of co-authored papers of any single European actor in the network. But it is individual initiative that helps intensify such ties.

The analysis also suggests many authors with a European affiliation can be assumed to be Chinese nationals studying or pursuing doctoral or postdoctoral research in Europe.

Military links

Unlike the UK where a number of universities are involved in collaborative research in AI with China, in Germany, Technical University of Munich has ‘outsized importance’ in research networks with Chinese partners, according to the MERICS report. The university is involved in seven of the top 10 bilateral links between German and Chinese institutions.

The report notes some of China’s leading AI firms and even academic labs have developed “symbiotic relationships” with public security organs in China.

For example, in 2022, a researcher affiliated with the Bundeswehr University Munich, a research institute of the German military, co-authored a study with colleagues from the PLA, according to the MERICS report.

The MERICS research also found “significant ties” with the so-called ‘Seven Sons’ of National Defense Xidian University, and the National University of Defense Technology – the top research and education institution of the Chinese military.

The ‘Seven Sons’ are nominally civilian institutions supervised by the Ministry of Industry and Information Technology but are a major source of research and personnel for China’s military and defence-industrial base. Xidian’s computer science research supports several Chinese defence and intelligence projects.

Identifying risky entanglements with nominally civilian universities and labs is less straightforward since many conduct significant amounts of defence- or security-relevant AI research, Arcesati notes. For example, one of China’s leading institutions for AI research, Tsinghua University, Beijing, carries out important military AI research and trains PLA students in computer science.

“If we want to make sure that there are no cases of knowledge or technology to undesirable end users in China, such as military-linked or security-linked end users or the military and state-owned military apparatus itself, then we really need to look at how to preserve research security and knowledge security in these dynamic dual-use (civilian-military) fields,” said Arcesati.

She noted that in Germany, there is now much more scrutiny in sensitive technology fields but that a lot of collaborations “don’t violate any law. In many cases export controls don’t even apply”.

“AI is not really well covered beyond hardware tech restrictions,” Arcesati pointed out, adding that export controls are not equipped to deal with emerging technologies such as AI Brain. “So in many cases universities cannot stop such collaborations from taking place even though some may involve security or surveillance applications,” she stated.