EU/China Relations, AI and the Realities of Technology Competition: An interview with Rogier Creemers

You may listen to this interview from HERE.  

Rogier Creemers is a Lecturer in Modern Chinese Studies at Leiden University in The Netherlands. With a background in Sinology and International Relations, and a PhD in Law, his research focuses on Chinese domestic digital technology policy, as well as China’s growing importance in global digital affairs. He is the principal investigator of the NWO Vidi Project “The Smart State: Big Data, Artificial Intelligence and the Law in China”. For the Leiden Asia Centre, he directs a project on China and global cybersecurity, funded by the Dutch Ministry of Foreign Affairs. He is also a co-founder of DigiChina, a joint initiative with Stanford University and New America.

Q1. Please introduce yourself and tell us about your current interests.

Perhaps the best way to do so is provide a little intellectual autobiography. I grew up in a small town in Belgium and decided to take a course in Sinology at university, largely out of an interest in the language, culture and history of a country so very different from mine. After graduation, I had the opportunity to spend a year at Peking University, where I discovered that the China of the present is at least as interesting as the China of the past, and that it would have a significant impact on the globalizing world of the mid-2000s. I subsequently took another degree course in international relations, focusing on China’s role in international economic organizations. That, in turn, led me to a PhD position at Maastricht University, where I analysed the relationship between WTO rules on intellectual property and Chinese practices, particularly in the realm of copyright. My conclusion was that copyright law in China is not only a market mechanism, but is also embedded within Chinese media regulation practices.

I then was appointed to a post-doc in Oxford University, to study Chinese media regulation more deeply. This was in 2011, when the number of Internet users in China was skyrocketing, and it thus became impossible to say anything meaningful about Chinese media without including the online sphere. I rapidly found that the Internet in China was about far more than online content, it also covered the emergence of China as a hardware manufacturer and innovator, of Chinese companies as bigger players in online services and platform, and the Chinese government as an ever more ambitious participant in global digital governance processes. This involves the two main thrusts of my current research: how does the Chinese government attempt to harness the capabilities of digital technologies for its own development programme, and how does it impact global digital governance and economic processes.

Q2. What is the European position on technology cooperation/competition with China and how do you see it evolving in the coming years? How does the EU balance the various interests of its member states to achieve a cohesive strategy that can be implemented effectively across countries with very different enforcement capabilities and self-interests.

I think the European position has changed very significantly, due to several successive events that led to a realization among European decision makers that the existing relationship with China was not as much to Europe’s benefit as first believed. A first was the “Kuka shock”, when German robotics manufacturer Kuka was taken over by Midea, a Chinese firm, in 2016. Kuka was seen as an important player in the “Industrie 4.0” initiative, aimed at modernizing the German manufacturing sector, and a symbol of engineering excellence. The fact that “China” could just purchase a crown jewel like that, without any meaningful way to review or even block such an investment, led to a far more critical stance about China’s growing ambition in Europe. A second was the Covid-19 pandemic. On the one hand, this revealed how much our manufacturing capability in Europe was hollowed out: we couldn’t even make our own face masks. On the other hand, it demonstrated the vulnerability of our high-consuming way of life to sudden shocks. Lastly, there is the Ukraine war, which caused many in Europe to reflect on our dependencies on a country that remains largely supportive of Russia. The problem is that the exposure of different European states to China, as well as the priority they devote to it in their own capitals, is highly divergent. We are therefore just at the beginning of a policy cycle where Europe needs to move beyond the clever-sounding but not-very-useful triad of “partner, rival and competitor” towards a clearer position of which China-related policy objectives are prioritized, and how the costs of attaining them will be borne.

Q3. The EU is seeking to protect key technologies and industries from China but at the same time it needs Chinese technology (solar, batteries, etc.). It also wants to ensure market access for its own companies, so how do policy makers thread this needle without harming overall trade relations?

For the moment, actual “decoupling” from China is relatively limited, with the obvious exception of the semiconductor sector. The difficulty is that Europe is trying to protect, ultimately, its high-end position in the pecking order that is the global distribution of income against a player that has decided it wants to reach into exactly those commanding heights. The EV sector already demonstrates China’s ability to do so in a highly innovative way. I’m not quite sure it will be possible to continue having our cake and eating it, particularly as we seem to be moving in the direction of a kind of industrial policy that isn’t a million miles removed from the Chinese approach. And this is an issue that will not just play itself out in the bilateral relationship, but also in third countries, where European and Chinese companies will compete increasingly for markets.

Q4. Looking at a specific example, what are your thoughts on ASML? How does a firm like that navigate the current geopolitical climate where it is under pressure to sacrifice some of its China business. Does ASML have to align with the Americans, or can it chart its own course? Is there a risk that by withholding critical technologies from the Chinese, you just give them the incentive and motivation to move up the value chain and eventually become formidable competitors?

Not that long ago, companies like ASML preferred to see themselves as a sort of techno-corporate Switzerland. Actually, a mere few years ago, the semiconductor industry largely operated under the radar of public and political view. The geopoliticization of chips has, obviously, changed this. Governments, analysts and businesses are all on a fascinating journey of discovery of how incredibly complex the value chain is, and particularly Washington has been seeking for critical bottleneck technologies where it can deprive China’s chip ecosystem of the oxygen it needs for further development. It is instructive to see how the screws on ASML have been tightened. First, Washington negotiated with the Dutch government to prevent the export of the most advanced EUV machines to China. Then, the 2022 sanctions still required Washington to work with the Hague on export controls over the less-advanced DUV machines but made it difficult for ASML’s US staff to work on China-related projects. Subsequently, Washington pressured the Hague to implement DUV sanctions earlier than the originally planned date, and the 2023 tightening of sanctions essentially introduced a rule that Washington could decide directly on ASML exports, without intervention from the Netherlands. ASML’s significant interests in the US, and its dependence on US technologies, suppliers and talents mean that it is difficult for the company to act autonomously.

Of course, there is a risk that this will actually accelerate Chinese technological progress. As long as Chinese equipment manufacturers had reliable access to high-end chips, usually produced by TSMC, the incentive for them to support China’s semiconductor industry was small. They could just go for the standard option. Now, companies such as Huawei no longer have access to these chips, and many other Chinese firms that aren’t yet sanctioned operate under the assumption that they might become so in the future. This means the existence of a competitive Chinese chip industry is of significant interest and importance to them. That doesn’t mean reproducing these capabilities will be easy: it has taken companies such as ASML and TSMC decades and billions to get to the current point. However, Chinese chip firms may also discover alternative technical paths, and there is no doubting the political will and the resources that the Chinese state is willing to make available.

Q5. The Russia/Ukraine war is a core issue for the Europeans and China’s position puts it at odds with the EU. What can the EU realistically expect of the Chinese in terms of its strong relationship with Russia? Can the EU potentially use this to its own advantage by encouraging China to employ its leverage (if it truly has any) on Russia to positive effect, or do you see this war remaining a key barrier to improved relations between the EU and China?

In short: yes. There has been little or no signal from the Chinese side that they understand the EU’s position vis-à-vis the Ukraine war, let alone that they might come to explicitly support Ukraine. Rather, China views the European position as largely an avatar for the US pursuit of hegemony. For the time being, China is seeking to exploit the crisis, amongst others for cheaper access to Russian oil and for extra export income through industrial support.

Q6. An elephant in the room is the potential election of Trump. His views on alliances and Europe tend to the transactional, and there are potentially serious divergences on Russia and Nato, as well as the possibility that the EU could find itself in the crosshairs of new US tariffs. While a serious breach in US/EU relations is unlikely, do EU policy makers need to consider a more neutral relationship with China as a potential counterbalance to a more unfriendly or insular US?

Certainly, EU policymakers need to be prepared to shoulder a far greater burden in the maintenance of their own well-being. As the refrain goes, Europe outsourced its energy provision to Russia, its manufacturing to China and its security to the US. The Russia situation is clear for all to see, China is moving up the value chain, and a Trump election would likely alter the US commitment to support security in Europe. The only alternative is that Europe steps up its own game, but this will be politically difficult: it will likely have an impact on Europe’s welfare state model, require a lot more military spending, and lead to a reduction of prosperity in Europe in general. Already, we see the populist right leveraging these themes for electoral success.

Q7. Shifting gears to AI, can you talk about the emerging legal and policy framework around AI and AI-generated content in Europe. There appear to be notable differences in approaches around things like copyright, ownership and intellectual property across the EU. It does raise the question of whether or not the EU has a real future in AI given the fragmentation.

The AI Act, which is likely to come into force by the summer, will already do a lot to harmonize Europe’s approach to AI, including in intellectual property-related issues. I believe the true crux of Europe’s AI future lies less in regulatory fragmentation, and more in the ability to create the financial support, market incentives and data resources that exist both in the US and in China. Europe has already missed the boat on much of the digital economy, and something similar might occur again in AI.

Q8. What are your thoughts on AI data scraping? How is it possible to regulate AI given that automatic data scraping tools do not differentiate between public and private data or data protected by copyright? Even datasets from non-profit groups have been found to contain scraped data from dubious sources so how do policy makers address accountability, consent, and attribution issues?

To a significant degree, this question echoes the assumption of the early Internet: the nature of technology would make regulation both impossible and undesirable. Since then, we have increasingly seen that the Internet can be regulated and, in very many cases, it is possible to make a strong argument that it should. That means certain things become more difficult or impossible, that is the price you pay. Obviously, in the realm of AI, this will be a protracted process, but a head of regulatory steam is already building, exactly because we have learned with the Internet what might happen if such issues aren’t addressed promptly.

Q9. More generally, as the web becomes overrun by generative AI content and the capabilities of Large Language Models (LLMs) increase exponentially, how can policy makers keep up with this pace of change and ensure human society can maintain trust in what they see and read online?

There are some solutions to the immediate issue of identifying AI-generated content, such as watermarking. However, the broader challenge is coming to understand how the nature of information production and dissemination has changed. All of our ideas, concepts and rules around content emerged in an era where content was expensive, and therefore its creation needed to be encouraged. That is, for instance, why we have a copyright term of life plus seventy years. However, we are no longer operating in an environment of scarcity, but of abundance. The amount of information out there is functionally infinite.

A major challenge will be how we reform our approach to content governance in a way that safeguards the values of openness and freedom we cherish, but also recognizes that the abundance of information requires new approaches regarding, for instance, ownership of major content platforms or outlets, transparency and identification of users, and the responsibility of gatekeepers, search engines and other forms of content selection. AI turbocharges that need and brings greater urgency to policymakers. I expect that within the next decade, we will see far more action to develop reliable online identity models, more rules to ensure a 21^st century version of public broadcasting rules and ethics in the online sphere, and requirements for clear identification around AI-generated content. At the same time, we seem to be going through a bit of a hype cycle with generative AI, and I’m also confident that much of the current doomsaying will turn out to be unfounded.

Q10. As a major tech player in its own right, what role do you see China playing on the global stage in terms of setting digital standards and helping shape the future of data policy and emerging areas like AI? Globally, is there much alignment between what China is doing and what is happening in Europe or the US?

Usually, standards in the digital realm serve the purpose of interoperability and ensuring that all kinds of different hardware and software can work together. China is seeking to grow its presence in global standards. European governments often interpret this as a power grab, but the reality is a little more prosaic. Digital standards are IP heavy, and inclusion of patents into a standard can generate significant income streams – very important for a country that still has to quadruple its per-capita GDP to achieve the same levels of prosperity as Western Europe. In the technical realm, it seems that Chinese companies are becoming ever more successful in this regard – which is not really a surprise given their rapid growth in capabilities.

Where it comes to governance, however, there is little or no interaction between China and Western states. To be sure, domestic measures often resemble each other. Chinese data protection legislation has obviously been inspired by the GDPR, for instance. But I do not see this leading to the emergence of a comprehensive global governance regime in the foreseeable future.

Q11. Looking at China, how do you view the “tech rectification” of recent years? What was the objective and where are things today?

For those readers suffering from insomnia, I am happy to refer to a working paper I wrote on the subject. In short, I believe the rectification served several different policy goals: maintaining stability in the financial system, regulating market competition and data use, addressing emerging social concerns such as labour conditions and child protection, supporting China’s new industrial policy, and limiting exposure to geopolitical risks. This is taking place in a context where China is moving to a new socio-economic development paradigm, which focuses less on outright growth. Instead, it takes a more selective approach to GDP (focused on industrial innovation), rejects financialization (which has an impact on fintech), and seeks a greater balance between the powerful platform companies on the one hand, and users as well as on-platform operators on the other. Platform companies are expected to behave as good corporate citizens.

There is little indication that the Chinese government regrets what it has done: the new regulatory status quo is here to stay. There is, however, a recognition that this took place in a disruptive manner, and this should not happen again. Recently, new draft rules on online gaming would have severely limited the earning potential of online gaming firms. After protests from these firms, the senior responsible official was dismissed. The very onerous data export regulations have been softened somewhat as well. So, there will be a difference of tone, if less one of substance.

Q12. There is a view that the Chinese government increasingly sees data as a factor of production and a key to future economic growth. How does the Chinese State (or any government) balance the competing objectives of maintaining control over sensitive data, ensuring privacy is respected, but also safeguarding that data are available for all parties to use, and not simply monopolized by a few large technology companies with the resources to dominate?

This will be the task of the new National Data Administration. Hitherto, the focus in data governance primarily was on security, with relatively little attention going to the developmental side of things. The NDA is studying the development of legislation and regulation to support the development of data markets, including in areas such as property rights, pricing and insurance. It is also exploring making more governmental data available for private sector use. The rectification shows that the government is aware of the risk that a few large platforms dominate this space, but it must balance that risk with the fact that the size of these platforms also mean they have the resources necessary to run large data centres and complex AI models. A last thing to bear in mind is the development of public compute infrastructure, which should reduce the cost of running AI models for start-ups and SMEs.

Q13. You are involved in quite a few multilateral events. What is it like being a Chinese speaker at these events, and what are your observations on the Chinese attendees? What do you see as the strengths and weaknesses from the Chinese on the diplomatic front? Do you see improvement from the Chinese side on “telling China’s story well”, as this is a key government objective?

I think the biggest challenge Chinese diplomats and participants in multilateral events have to face is that indeed, “telling China’s story well” essentially means sending a message that is agreeable to the central leadership in Beijing, but that does not necessarily gel well with overseas audiences (although we Europeans should not flatter ourselves too much about the attraction of our story either). Because it is a China-centric story, there is little space or ownership potential for other countries to write themselves into that story. On the other hand, particularly for developing countries, China is both an example to be emulated, and a supplier of high-quality, low-cost infrastructure.

Q14. As someone with a deep interest in Chinese and European affairs, do you find it difficult to straddle both worlds? Are you finding collaboration with Chinese colleagues still possible, and can you get access to the Chinese information and data you need to support your work? On the European side, do you find views on China to be nuanced and informed?

Things have indeed been getting more difficult, not in the least because of the pandemic and the aftermath. Sometimes, it’s little things: it’s still impossible to take the underground in Beijing for foreigners, because a Chinese ID is necessary to get a ticket out of the machine. Most of my work is based on policy and regulatory texts, and these tend to be easily available in open source. However, some databases, such as the Wind platform for business information, have become less easily accessible. I am pleased to say that my Chinese colleagues are as energetic and motivated as ever, particularly as we not only work together on research, but also on maintaining expert dialogues between the Chinese and European sides. This is especially necessary as knowledge about China in the EU is lamentable. We simply have not seriously invested in structural research capability, with the result that we are now scrambling to get even basic information about what is going on. There is much room for improvement here.

Q15. Please share any favorite books, blogs, podcasts or other resources that readers could use to improve their understanding of EU-China relations, tech policy, AI or other relevant topics.  

I’m going to shamelessly use this opportunity to highlight a few works of which I have been a part. I was fortunate enough to bring together a wonderful group of people and write a book about multiple aspects of China’s emergence as a tech player. Thanks to the Dutch MFA, our funder, the digital version is available free of charge.

Also, I am a member of Digital Power China a research project led by Tim Ruhlig at the DGAP in Berlin, which assesses China’s impact across a range of technologies, with teams composed of both policy and tech experts. We have already published two reports, and third report is going to come out in May.

On tech, some great writers include Martin Chorzempa (on fintech), Matt Sheehan (on AI), Kendra Schaefer (on tech in general), John Lee (on semiconductors). On EU-China relations, Alicia Garcia Herrero is very good on the economic side, and Frans-Paul van der Putten on the politics.

If you have time for one China podcast, let it be the Sinica podcast, run by Kaiser Kuo. It covers the country in such breadth and depth, it’s a masterclass in and of itself. Other good podcasts include Pekingology (CSIS), Chinese whispers (Spectator) and the Little Red Podcast, run by some excellent people at ANU in Canberra.

If you or your company has a due diligence requirement or needs assistance on a fraud issue, corporate investigation, risk advisory, or other related matters, please contact us at info@kalavinkaadvisors.com or +852 2196 2727