TITLE The race to Artificial Intelligence supremacy in unfair conditions Subtitle A study on how privacy rig

(1)

The race to Artificial Intelligence supremacy in unfair conditions

A study on how privacy rights act as a barrier to the European Union’s goal of being a leader in AI in the face of Chinese competition.

Bachelor’s Thesis

Faculty of Economics and Business BSc Business Administration

Student name: Yannick Andreas den Boer Student number: 11442409

Supervisor: Erik Dirksen MSc Coordinator: Dr Michele Piazzai Ethics approval: EC 20200721100712

(2)

Statement of Originality

This document is written by Student Yannick Andreas den Boer, who takes full responsibility for its contents. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

Current research in the field of Artificial Intelligence and business focuses on the economic impact of its introduction. In this study, the author took the perspective of how privacy rights affect the European Union’s ability to compete with China in the field of Artificial intelligence. Using interviews and a content research analysis, several insights were developed to attempt to explain how the competitive field is shaped when looking at competition from a privacy-centric point of view. It was found that while privacy rights play a role, access to data is a more important factor.

At the same time, the lack of a common goal in the European Union in developing its Artificial

Intelligence industry is disadvantageous compared to China’s more unified vision. Lastly, the definition of success in developing Artificial Intelligence cannot be measured solely in economic terms.

(4)

1. Introduction

In the Chinese city of Sanmenxia, bordering the Yellow River, local law enforcement began trialing a new facial recognition technology in 2019. In one month, residents were screened half a million times to identify whether they were Uighurs, a persecuted Muslim minority (Mozur, 2019). A few months later, the Swedish Data Protection Agency fined a municipality for “tracking 22 students over three weeks and detecting when each pupil entered a classroom” (School’s facial recognition checks

lead to fine, 2019).

Much has been said in recent decades about the growth of China as an economical behemoth that gained technological advantages by copying successful Western companies. While this was true for a long time – and is still happening in certain industries – China’s lead on innovation in technology has grown considerably and is a competitor in areas previously dominated by Japan and the United States. While technology encompasses many elements, an area where they have been focused on heavily is automation and using artificial intelligence to help bolster not only their manufacturing abilities but also consumer led areas such as e-commerce and transportation systems (Deloitte China, 2019). This is also seen by their commitment to focusing on Artificial Intelligence in their Made in China 2025 program. The country is now seen as a competitor to the United States, which is widely considered the leader in Artificial Intelligence. This growth is also seen as a concern for the European Union as some leading experts doubt the EU will be able to compete with China on AI in the future (Castro et al., 2019).

The impact of looking at China in a belittling manner has led to a lack of strong initiatives in other parts of the world where private corporations (in the case of the United States) and a mix of private and public-private partnerships (in the case of the European Union) are most common. China’s rise to economic power and its form of governance allow it not only to invest huge amounts of money in research, but also to form long-term plans to which Chinese companies must adhere, such as the Made in China 2025 plan (Li, 2018).

(6)

The EU itself acknowledges: “Investment levels for Artificial Intelligence in the EU are low and fragmented, compared with other parts of the world such as the US and China” ("Artificial intelligence", 2018). This foreshadows both an issue and a challenge that need to be resolved with the utmost urgency if European countries want to avoid falling into a state of naivety and, consequently, technological stagnation.

This brings us to the question: Will the lack of stringent privacy rights in China irrevocably damage the European Union’s ability to compete with China’s rise in Artificial Intelligence? The United Kingdom has been left out as they are no longer formally in the European Union at the time of writing.

(7)

2. Literature Review

The topic being researched is focused on how the European Union can still compete with China on the front of Artificial Intelligence given the immense differences in both blocs and a lack of a clarified vision in the EU. This is an area which has been stated as being important to the EU and has received attention in the past few years. To discover and understand the key topics at hand, several areas of existing literature are discussed. They are related to how technology (and especially AI) is changing the business landscape, how innovation management theories regarding governmental subsidies and efforts are able to help innovation in AI, what the state of AI in the EU and China currently is and how ethics and privacy rights come into play.

2.1 Integration of Artificial Intelligence in Business

The technological landscape of the world has changed drastically in recent years with the introduction of Information Technology. The rise of IT in everyday applications across multiple fields has been called the fourth industrial revolution, represented by “hyper automation and hyper connectivity based on artificial intelligence (AI), big data, robotics, and Internet of things” (Park, 2017).

The introduction of AI has had a fundamental impact on businesses in many industries, and it could have a colossal effect on countries’ economies. This is partly due to the fact that AI changes the way people work with technology and the degree to which technology functions on its own. It also means employees will have to adapt to working with technology even more. Artificial Intelligence has the potential to boost productivity, especially as “36 of 37 advanced economies had slower productivity growth during 2006–2016 (1% average growth) than during 1996–2006 (2.7% average growth)” (Furman & Seamans, 2019).

There is no clear-cut definition of Artificial Intelligence. This is because the meaning of AI differs depending on the goal of its implementation. However, the European Commission proposed an

(8)

environment and taking actions – with some degree of autonomy – to achieve specific goals.” As there are many different ways to describe AI, another proposition by Kok et al. (2009) is to classify them into four categories: “Systems that can think like humans, systems that act like humans, systems that think rationally and systems that act rationally.”

Using technology such as Artificial Intelligence has become increasingly prevalent in companies seeking to maintain a competitive edge. A McKinsey Global Institute report by Bughin et al. (2017) highlights the rise and relevance of AI, estimating that firms spent between $18 billion and $27 billion on internal corporate investment in projects related to AI in 2016. A 2019 survey of 10 industries found that 30% of organizations were conducting AI pilots and 47% were using at least one AI capability in their standard processes – up from just 20% in 2017. More than 70% also expected AI investment to increase significantly in the coming years (Wladawsky-Berger, 2020).

This rapid growth has been apparent throughout all sectors and not only among technology companies. Active research in Artificial Intelligence has been gaining steadily in many fields, including business, finance, accounting, engineering education, science and medicine (Al Sheibani, Cheung & Messom, 2018). For example, Google had two deep learning projects (high-level Artificial Intelligence). in 2012. By 2016, it had more than 1,000 (Makridakis, 2017).

One reason for this rapid growth is the need to handle and extract information from the vast amount of data generated daily. As of 2012, more data crosses the internet per second than was stored in the entire internet in 1992. At the same time, around 2.5 exabytes of data are created every day, which is equivalent to 2,500 petabytes. Three-fourths of this data is unstructured and it is doubling every 40 months. To put this in perspective, Walmart collects more than 2.5 petabytes (2,500 terabytes) of data each hour from its customer transactions, which is equivalent to 50 million filing cabinets’ worth of text. Capturing the value of this data is thus not just important – it is crucial for businesses to succeed

(9)

in this new era. Most importantly, Artificial Intelligence can play a key role in translating this data into meaningful insights (McAfee & Brynjolfsson, 2012).

It is logical that countries consider the ability of their economy to harness the power of AI as paramount. According to PwC (2016), AI has the potential to contribute $15.7 trillion to the global economy by 2030. Understandably, governments are pushing for their businesses to research AI and to develop new applications to help succeed in an area with such potential. At the same time, countries around the world are investing significant amounts of money in Artificial Intelligence – both because it will have a major impact on the economy and because they want a role in deciding how AI is

implemented. The latter is important because countries have different views on ethics and therefore different views on how AI should be used (Anderson & Anderson, 2011; Scholtens & Dam, 2007).

2.2 Universities and Government-Funded Research Bolster Innovation

An important factor in the area of innovation management is how companies can develop and further their innovative cycles. Government- and university-funded research that trickles down to the private sector has been cited as a propagator of this (Schilling, 2017). Cohen and Levinthal (1990) found that many companies reported that research from public and nonprofit institutions enabled them to develop innovations that they otherwise could not have. Public funding also has a positive effect on a company’s R&D expenditure. According to Afcha & León López (2014), public funding, regardless the level of government grants, positively stimulates R&D expenditure.

China’s recent 2025 roadmap identifies 10 industries in which it wants to become a world leader by 2025. Two of these – information technology and automated machine tools & robotics – include Artificial Intelligence as keys in their success ("Building a world manufacturing power," 2017). On the other hand, countries in the European Union each have different economies, currencies and

(10)

The amount of research on AI is also important, because it will allow AI to be used in a new way on applications – which is the most difficult part of making Artificial Intelligence useful . According to Castro et al. (2019), the European Union has more AI and top AI researchers (using the H-index, which measures the productivity and influence of researchers) than both China and the United States in absolute terms (the US has a higher number per capita). However, the EU seems to be failing to capture the value of its own research because of fragmented markets, strict competition regulations, a lack of large conglomerates (none of the top 10 R&D spending firms in the world are European) and the fact that there aren’t any European equivalents to tech giants such as Alphabet, Baidu and Amazon (IC Insights, 2020).

The differences in spending are also stark. China announced plans to develop a $2.1 billion AI-centric technology park in 2018. The Tianjin municipality has announced a $16 billion AI fund to aid its goal of becoming a world-leading AI innovator by 2030 and plans to invest “tens of billions more” (Chen, 2018, Knight, 2019). In the EU, the largest economies, France and Germany, have announced plans to invest €1.5 billion and €3 billion on research, respectively (Loucks et al., 2020). While the European Commission did call for a €20 billion cash injection for Artificial Intelligence research, no large project has materialized, and EU countries are taking national approaches. The new multiannual financial framework (2021-2027) may have a larger budget available for AI-specific research (Rankin, 2018).

In the field of innovation management, a country’s spending on Research and Development has been shown to positively influence the ability of its companies to harness new technologies and further develop them. Gittleman & Wolff (1995) mentioned that “a voluminous literature has demonstrated that R&D makes an important contribution to growth at the firm, industry and national level”. Figure 1 shows the evolution of public and private spending on R&D based on gross domestic product in the world’s four largest economies. As can be seen, China outspent the EU for the first time in 2015.

(11)

Figure 1: Gross domestic expenditure on R&D, 2006-2017 (%, relative to GDP)

Note: This figure includes the United Kingdom as a benefactor, though this was nullified when the country exited the European

Union. Copyright Eurostat R&D Statistics explained (rd_e_gerdtot)

Another interesting and more concrete difference is the absolute change in Research and Development expenditure for the EU, China and the United States. Figure 2 shows that while all countries have increased their spending on R&D in recent years, there are remarkable changes in absolute terms. The spending based on GDP illustrated in Figure 1 doesn’t show the full picture, as gross domestic product is not a relative comparison.

In 2018, real GDP was $13.6 trillion in China and $15.9 trillion in the European Union (World Bank, 2020). Between 2012 and 2015, the European Union increased R&D spending by €27.2 billion – less than a third of the additional €87 billion that China spent (based on Purchasing Power Standards). Purchasing power standard is an artificial common reference currency unit used in the European Union that eliminates the differences of price levels between countries ("Definition - Purchasing power standard | Insee", 2016). This is a significant difference, given that the European Union has a larger GDP

(12)

than China, which has spent more in recent years than both the European Union and the United States. China has also been funding R&D more than the European Union since 2013.

Figure 2: Change in R&D Expenditure between 2012-2015 in Euros

Note: Data gathered from Intramural R&D expenditure (GERD) by sectors of performance from Eurostat

On a macroeconomic level, the large size of China’s economy and its ability to centralize

planning in different sectors gives it an ability to funnel funds to its desired sectors of research. While all European Union members did sign an Artificial Intelligence cooperation declaration, it still doesn’t compare to China’s investments and ability to centralize planning and steer its private companies in its desired direction ("EU Member States sign up to cooperate on Artificial Intelligence - European

Commission", 2018). Also notable is the population difference, China has a population almost three times larger than the European Union ("Living in the EU | European Union", 2020; United Nations 2019).

(13)

This can have a longer-term effect of China being able to educate more people in the fields of Artificial Intelligence than the EU possibly can.

2.3 The state of AI in China and the European Union

In current times, one of the most talked about area in technology is the rise of Artificial intelligence (Deloitte, 2020). This new area of Information Technology is so revered due to its ability to change the way humans interact with computers and the amount of labor which can eventually be automated and done by computers and machines. At this time, the leading country in AI is the United States, which is strengthened by its highly ranked universities and research programs in the domain of Artificial Intelligence. China trails as second, with reports that its only advantage at this point is access to data (Ding, 2018). Another reason cited for why the European Union is lagging behind both China and the United states is that the EU has “a weak industry, in part because venture capital funding of AI startups in the US and China dwarfs Europe” (Ding, 2018). Unfortunately, it is difficult to find

quantitative data and metrics on how China and the EU are specifically investing in all areas of AI and how it is progressing.

As can be seen from Figure 3, there is also a clear difference in thinking about the risks of AI compared to its payoff between European Union countries and China.

(14)

Figure 3: Concerns about potential AI risks compared to the confidence in the ability to address them across several countries.

Note: This graph was adapted from Deloitte State of AI in the Enterprise survey, 2nd_{edition, 2018.}

2.4 Contrasting Privacy Laws and Ethics

Another interesting aspect is the impact of the legal systems in the EU and China. Stringent privacy rights in the EU can be seen as being a barrier for the development of AI. China, on the other hand, doesn’t have a strict framework with clear definitions of individual rights. And when there are definitions, they often only focus on how companies deal with people’s data and disregard the status of privacy rights the government needs to respect. (Wang, 2012). China has an authoritarian government, where there is no distinct separation of powers between government branches. While a country in the EU would need to seek lawful court orders to gather data on its citizens, China has no such need.

China’s Social Credit System, which came into effect in 2014, exemplifies this reality. This system creates a unified record system for individuals, business and the government to be tracked and

evaluated for trustworthiness. This score can later be used as a “reward or punishment mechanism”. Chinese government agencies do not need to obtain the consent of data subjects to collect financial data, which is a stark contrast to the EU’s General Data Protection Regulation. (Chen & Cheung, 2017).

(15)

An example showing the scope of data use in China by the government is facial-recognition technology, which has a direct effect on the development of Artificial Intelligence. In 2017, it was reported that China had more than 176 million surveillance cameras. These devices use deep-learning algorithms to “assess the number and density of people in the frame, individuals’ gender, and the characteristics of clothing and vehicles”. The goal was to have 100% coverage in designated areas such as public spaces and residential communities by 2020 (Qiang, 2019). Some facial recognition cameras in China are used to identify whether people belong to a minority such as the Uighurs, who are oppressed (Daly, 2019).

There is a large-scale discussion today about the ethics of using AI and how data can be used to further it. The debate doesn’t only surround the ethics of using AI – it also concerns how ethics are defined. A study of how countries and businesses are dealing with it by looking at how many documents were released providing guidelines for the use of AI revealed that the United States had the highest number, with 21, followed by the EU’s 19. Interestingly, there was no data for China, though language barriers may have been partially at fault (Jobin et al, 2019).

However, the Beijing Academy of Artificial Intelligence did recently publish guidelines called the Beijing AI Principles, which made some landmark changes. Backed by the Chinese Community Party, the release of these guidelines was seen as a positive move because it allows for a more open discussion on ethics and human rights (Knight, 2019). China’s involvement with the World Economic Forum’s AI principles sent a positive signal to other countries that it is willing to regulate its use of data further, though critics still are wary due to the fact the ruling party’s tight control over domestic businesses hasn’t changed (Knight, 2019).

Another aspect of ethics is how it can have an effect on the type and amount of innovation companies output. mall scale studies have also shown that ethics can have an effect on innovation due to some factors such as knowledge-sharing. A study by Awan & Akram (2012) showed that “the amount

(16)

of knowledge sharing has a moderating effect on innovation”. At the same time, it was also shown that when organizations have the “right” morals, innovations can be even greater than if there was a lack of morals or ethics in the organization (Nunn & Avella, 2015). Thus, there is cause to believe ethics and morals have an effect on innovation in both the EU and China.

The aforementioned theoretical insights on the topic of China and its rise in Artificial Intelligence signal a clear trend that may have important consequences for business in the European Union which should be addressed. While the current literature looks at AI’s ethical and business implications and the change it will bring in the future, there is a lack of knowledge in how different countries are each dealing with Artificial Intelligence. There is also a gap in knowledge on comparative papers between China and other AI industries, as many of the major and most cited papers aren’t focused on comparing China and how it is adapting AI compared to other countries.

Equally important is the lack of academic papers examining how the different privacy protection schemes in the EU and China lead to unequal performances. This question is relevant to both academics and institutions in the European Union to better assess to what extent the EU is willing and able to change its stance on privacy rights to achieve a leading position in Artificial Intelligence.

(17)

3. Methods

3.1 Data Collection

To investigate how privacy rights may play a role in the European Union’s future competitive position in Artificial Intelligence, data was collected using two semi-structured interviews. Semi-structured interviews were chosen to deeply explore the developments in the interviewees’ fields and to understand their experiences. One interviewee was an expert in privacy law and the other an expert in Artificial Intelligence. Both are Dutch and living in the Netherlands. They were chosen based on purposive sampling due to their knowledge in their respective fields, their interest in privacy and technology, and their involvement with European Union privacy and AI projects.

The interviewees were contacted by e-mail and LinkedIn. The interviews were carried out online through Zoom conferencing software and lasted about 50 minutes each. Several questions were used to gather concrete answers from the interviewees, after which they were asked to elaborate on their answers to gather more insightful data. The advantage of using interviews in this research is that experts in the fields of AI and privacy law are up to date with current developments and are able to elaborate on answers. By using interviews, it was also possible to develop more theories and ideas to answer how privacy laws are impacting the European Union’s competitive position based on these discussions.

3.2 Analytical Procedure

Prior to the interviews, guides were made to ensure a clear and consistent structure when interviewing. A funnel technique was also used, meaning the questions progressed from general topics to more specific and comprehensive questions. By doing so, the participants were able to become accustomed and more comfortable in answering questions, allowing for better analysis post-interview. The analysis was performed using a manifest content analysis approach. This method was chosen for its ability to find key messages and themes from the gathered data. By doing so, it complemented the

(18)

inductive approach of building theory rather than testing it. The coding of the transcripts was done inductively and using an open coding approach. The main themes and categories were then extracted and used for theory building.

(19)

4. Results

The aim of the research was to determine whether the lack of stringent privacy rights in China irrevocably damage the European Union’s ability to compete with China on Artificial Intelligence. The results are summarized in a table that is then further developed to provide more context to each theme and topic identified. Due to confidentiality, the interviewees were labelled as A and B for the privacy and AI expert, respectively.

4.1 Key Themes and Topics

Table 1 summarizes the most important findings from the research. Through analysis, three themes were found and are provided below with quotes from the interviews.

(20)

Table 1: Key Themes and Topics extracted from the interviews

Regulatory scheme Data needs Ethics and mentality

“In the Netherlands, the authorities are really strict. The government is therefore reluctant to start using data without a new law”

“Is it really necessary to collect all this data? Is it really being used or only 5%?”

“People in the Netherlands are critical of laws such as the spoedwet for the coronavirus” “I think they (companies) see this

(strict privacy regulation) negatively because they have to protect data. They also have to comply with all kinds of regulation”

“Companies in the EU will have less data because of the GDPR”

“Based on what I’ve read, I think Chinese people are more bound to listen to what experts say on a certain matter while here in the Netherlands, people will focus on their own opinion” ”in China, there’s not a clear

distinction of roles between the government and the company”

“If you don't have data, then you don't have algorithms most of the time. So, of course, it plays a role”

“Within the EU there are differences as well. There was a cookie legislation, in the NL it was quite strict but in others less so”

“the main reason why they (tech companies) are a little hesitant to enter the EU market, for example, is because they can get serious fines”

“you can't have a company, a private company, monopolize the data of people”

“I think that some countries in the EU may interpret the GDPR less strictly to help their (economy and companies) competitive position”

“If the government is going to propose that sort of stuff (laws allowing increased access to data for defense reasons), then basically everyone goes to the streets demonstrating”

4.2 Regulatory Scheme

One theme that quickly became apparent as being important was the regulatory scheme

surrounding privacy in the European Union. As interviewee A is a privacy lawyer and focuses on EU laws, she discussed how the European Union’s GDPR (General Data Protection Regulation) is interpreted and applied differently within the EU. She said that some countries, such as the Netherlands, apply and interpret the GDPR more strictly than others, such as Poland. This tied into the Netherlands being less willing to create new laws that are seen as invasive to people’s privacy, as their regulatory body and courts require extensive justification for such new legislation.

(21)

Another point that was discussed was how this regulatory scheme changes the competitive landscape. This expert said she had observed that some companies see privacy protection regulations as a negative thing, as they are required to do a lot more than before to protect the rights of their users, customers and other stakeholders. The necessity to do so also disadvantages companies compared to others that are regulated by other authorities or adhere to less strict privacy right regulations.

Regarding competition, interviewee A said it was difficult to judge China’s position in areas such as privacy guidelines, as it is difficult to distinguish between the roles of companies and the government. She also noted that many health companies are wary of entering the data commercialization space due to the sensitivity of data and the fact that the regulations are especially strict in this area within the EU.

Interviewee B touched on this same point by mentioning that some foreign (non-European Union) companies are hesitant to base themselves in the EU is because of the strict regulations that protect privacy rights and competition for both larger and smaller companies. He mentioned the example of Google being fined more than a billion euros for its role in an antitrust case. Interviewee B also predicted that data would become more of a “curse than a blessing” once regulations in countries like the United States become stricter and companies such as Google will have much explaining to do.

These ideas tie into the way the landscape in the European Union is shaped by its regulatory ties. These ties define how companies can go about gathering, storing, analyzing and commercializing data. Both interviewees agreed that they were important in shaping the future of Artificial Intelligence and its applications.

4.3 Views on Privacy

The second theme, views on privacy, relates to how different views on privacy have an effect on answering the research question. It became apparent during the interviews that both respondents were privacy rights advocates, and both mentioned they didn’t believe “handing in” these rights for economic welfare was a viable solution.

(22)

Interviewee A offered an example of privacy in the Netherlands, where there was a widespread discussion about a surveillance law that was put up to a referendum a couple of years ago. She found that it was too intrusive in many areas and said it didn’t make sense for the government to collect such data when maybe only 5% of it is actually required. Most Dutch people voted against the measure and the government later limited the scope of the law. On the other hand, regarding the GPDR discussion, she also mentioned that it is a given that EU companies will have less data than others that enjoy less strict privacy laws.

In the same gist, interviewee B solidified the importance of the need for data when he mentioned that for most applications, data is needed to create suitable solutions that often require algorithms trained with data. When it came to the differing views on privacy, he also insisted that a private company should not be able to monopolize the data of people. He discussed the amount of data some large companies such as Google have and how it can create inequalities in the future as a handful of companies will own data which are not only sensitive to people but can also be used to overpower their competitors due to which they are currently facing anti-trust lawsuits. He also mentioned that at times, some governments lack access to the correct experts in certain fields to help create legislation to reign in companies that own vast amounts of data. The theme surrounding data needs and what is adequate for companies gathering data for potential applications was tightly linked to ethics and mentalities as the decision surrounding data needs are guided by the latter.

4.4 Ethics mentalities

The third and final main theme discovered was ethics. According to both respondents, the ethics defining societies and their views on authority play a defining role in how privacy rights come into play. Interviewee B mentioned that based on his knowledge, it is more accepted and habitual in Chinese culture to be collectivist and adhere to higher authority. At the same time, Dutch citizens would be less likely to do so.

(23)

When an example was given of a hypothetical situation where the government would create new laws infringing on privacy rights, respondent B believed it wouldn’t pass even for the reason of national security and that people would demonstrate to show their opposition. Interviewee A gave the example of the coronavirus emergency law, which has been a point of discussion in politics where certain countries created an application to trace people through their location data. She mentioned that this had been heavily criticized in the Netherlands because Dutch citizens are not convinced and they want evidence it works before they give up their rights, even in the midst of a pandemic.

On another note, the difference in ethics is also apparent in the European Union as countries take different stances on certain issues. An example was the cookie law that came into force several years ago across the EU. As with many directives, government can interpret and enforce it in ways they see fit for their citizens. While the Netherlands had been strict about this, other countries in the EU didn’t implement strict regulations. At the same time, when asked whether she believed national governments may be less strict in certain areas to help domestic companies, interviewee A said she believed this may the case in some countries.

5. Discussion

Will the lack of stringent privacy rights in China irrevocably damage the European Union’s ability to compete with China’s rise in Artificial Intelligence? The research has indicated that the answer is no, as the issue at hand has a broader scope than foreseen as will be further explained. The main challenge for the European Union to match China’s rise in Artificial Intelligence is a fragmented union with different standards, regulations and goals. An important factor is the access to data that China has, and which is governed by a different set of rules than in the European Union. The research has led to three insights, namely:

(24)

Insight 2: The lack of a common goal in a union wanting to develop an industry will struggle comparing to a unified vision.

Insight 3: Success in the field of Artificial Intelligence cannot only be measured in economic terms.

5.1 Insight 1: Access to data is more important than privacy rights frameworks

During the discussions with both interviewees, it became apparent that there was a difficulty in assessing the full situation in China due to the government’s tight control on data and its willingness to share said data. However, there seems to be a consensus that the European Union’s mindset is to protect the privacy rights of its citizens without sacrificing their rights in order to improve their economic position.

While the focus of the research question at hand is whether the lack of strict privacy rights in China will have an effect on the EU’s competitive position, it appears that a more important aspect is the access to data. As mentioned in the literature review by Ding (2018), China can access more data than other countries due to its large population. However, this access to data has fewer limits than those in the European Union, regardless of the population size due to the fact that there isn’t a clear boundary between the government and companies. This is both due to the legal system as Wang (2012) observed, where the government can access private data from companies by bending laws in its favor. Access to data is crucial for the development of future applications of Artificial Intelligence as much data is required to build, train and test algorithms. The glaring contrast in the development of new AI

applications between China and the EU is important, as it reflects the ethics and mentalities that differ across cultures. As mentioned in the literature review, it is used in China regularly for facial recognition and other applications that are seen as intrusive, invasive and thus illegal in the European Union.

(25)

5.2 Insight 2: The lack of a common goal in the EU to develop its AI industry is a

disadvantage compared to the unified vision of China.

For all the merits of democracies, being able to point an economy into a direction the

government wishes seems to have its benefits as well. The European Union has taken a difficult path in Artificial Intelligence because of its plans and visions for the EU’s AI industry. While ideas are created and documents are signed by its members, many do not enact actual change due to national interest. On the legal front, privacy and technology law are interpreted and regulated differently, creating unfair competitive advantages within the union.

When comparing the EU’s strategy to the one in China, there is a clear difference in economic terms – both when looking at overall R&D spending as mentioned in Figure 2, but also in their

investments in AI, with China outspending the EU on most fronts (Chen, 2008, Knight, 2019, Loucks et al., 2020). This is a danger for the EU, as creating a competitive and fruitful industry requires

government funding and focus in the early stages. When the difference in spending on projects is coupled with the fact that EU companies receive a fraction of the venture capital their competitors in China get, this creates a hazardous situation for the future in line with Ding (2018)’s train of thought. Fortunately, the academic communities are used to cooperation and are heavily involved in AI research, which creates new opportunities for significant discoveries. It has become apparent that part of the reason some countries are currently leading in AI is not only due to funding, but also because they are founders of the computer age (in the case of the United States). However, as interviewee B also stated, academic focus and research on AI is seen as a very important factor for an economy to become an eventual leader, as this creates possibilities for experts in several AI fields to come together to combine their knowledge and develop new ways of thinking about Artificial Intelligence (EPRS | European Parliamentary Research Service, 2019).

(26)

5.3 Insight 3: Success in Artificial Intelligence cannot be measured only in economic

terms.

One thing that became apparent was that the focus of success for a country is often measured in economic terms. However, in the case of Artificial Intelligence, the definition of success goes further than monetary benefits. This is due to its vast potential, which spans almost every industry and influences and alters the daily lives of citizens. While some Chinese companies may deem their AI projects a success, as they have been bought by government agencies, others may say it is no success at all as they created applications that are used to monitor and target certain people.

This duality brings an important factor into play, which is the ethics of AI. While there are no widespread common guidelines on what should and shouldn’t be done with AI, some of the largest companies in the world have come together to create frameworks to avoid harmful applications of their research. At the same time, academics are also mainly focused on the theoretical side of AI and future applications, more so than creating unethical ways to use it (Hagendorff, 2020).

While in business, a project may reach success once it is able to generate a sustainable profit, for countries, success of an AI project may be measured by being able to lift people out of poverty, by decreasing fraud or perhaps by decreasing their CO2 footprint. While all three measures still have links to economic terms – which is logical considering that is the main unit of measurement in the world – they have effects far exceeding an increase in wealth or reduction in welfare payments. The EU seems to be heading in the right direction on the front of ethical use of Artificial Intelligence and can therefore be seen as successfully implementing AI for its benefit. While Artificial Intelligence idealists would disagree that profits should play a role at all, it is the role of an economy to generate more benefits than costs, and this is a discussion which should be held between the private, academic and public sector.

(27)

5.4 Significance of the results and limitations

The analysis lead to similar results as found in prior theory, however, there hasn’t been much research done on the specific question at hand. The insights evolving from this research are significant though open to interpretation as it is up to people to define what they see as more important. To some, the economic bottom line is all that matters while to others there is a broader issue at stake. The implications of this research can be of aid to policy makers who want to better their country’s economic standing as the role of AI is and will be a significant factor in the future of the world economy. Being able to harness the power in an ethical manner while understanding the competition faced will be a

discussion one should hold now before it is too late. This body of work adds a new perspective on the implementation of Artificial Intelligence across different countries and how norms, ethics, regulatory frameworks and fragmented markets may be a significant barrier to its success.

While the results from the two interviews are found to be significant, the study does have clear limitations. The study should be reproduced with participants from more countries in the European Union, especially ones which aren’t as economically sound as the Netherlands. There should also be participants from China who can offer a contrasting point of view to those already interviewed. At the same time, the number of participants should be enlarged to include a broader set of participants such as economists, businesspeople, and politicians as well. By doing so, there would be a clearer description of how the issue is seen on a more diverse scale.

(28)

6. Conclusion

During this research, the focal question being investigated was: Will the lack of stringent privacy rights in China irrevocably damage the European Union’s ability to compete with China’s rise in Artificial Intelligence? Based on the discussions and the theoretical insights, the question entails a large amount of context which cannot be uncovered in one study. Thus, the answer is no. The reason why the

European Union lags behind China in the field of Artificial intelligence isn’t only due to privacy rights but a multitude of reasons. However, there were three insights which may better help answer the question, namely that: access to data is more important than privacy rights frameworks, that the lack of a

common goal in the EU to develop its Ai industry is a disadvantage compared to China’s unified vision and lastly that success in artificial intelligence cannot only be measured in economic terms.

There are still some questions which remain unanswered such as which ethical norm will prevail and how the EU will create its budget for the next few years and how much it will focus on funding AI compared to previous years. In the end, the European Union must make a clear decision as it endeavors in creating and sustaining an industry rapidly changing and it must back up its goals with resources, and a framework for cooperation between its members.

(29)

7. References

Afcha, S., & León López, G. (2014). Public funding of R&D and its effect on the composition of business R&D expenditure. BRQ Business Research Quarterly, 17(1), 22-30. doi:

10.1016/j.cede.2013.01.001

Artificial intelligence. (2018). Retrieved 23 March 2020, from

https://ec.europa.eu/commission/news/artificial-intelligence-2018-dec-07_en

Awan, K. Z., & Akram, M. (2012). The relationship between Islamic work ethics and innovation capability and knowledge sharing plays moderation role. International Journal of Economics and

Management Sciences, 1(8), 34-48.

Building a world manufacturing power — Premier and Made in China 2025 strategy. (2017). Retrieved 9 June 2020, from

http://english.www.gov.cn/premier/news/2017/01/29/content_281475554068056.htm Bughin, J., Hazan, E., Ramaswamy, S., Chui, M., Allas, T., & Dahlström, P. et al. (2017). Artificial

Intelligence: The Next Digital Frontier? MGI Report, McKinsey Global Institute. Retrieved from

https://www.mckinsey.com/~/media/McKinsey/Industries/Advanced%20Electronics/Our%20Ins ights/How%20artificial%20intelligence%20can%20deliver%20real%20value%20to%20companies /MGI-Artificial-Intelligence-Discussion-paper.ashx

Castro, D., McLaughlin, M., & Chivot, E. (2019). Who Is Winning the AI Race: China, the EU or the United States?. Center for Data Innovation. Retrieved from http://www2.datainnovation.org/2019-china-eu-us-ai.pdf

Chen, Y., & Cheung, A. (2017). The Transparent Self Under Big Data Profiling: Privacy and Chinese Legislation on the Social Credit System. Journal Of Comparative Law, XII(2). doi:

(30)

Chen, Y. (2018). China's city of Tianjin to set up $16-billion artificial intelligence fund. Retrieved 22 May 2020, from https://www.reuters.com/article/us-china-ai-tianjin/chinas-city-of-tianjin-to-set-up-16-billion-artificial-intelligence-fund-idUSKCN1II0DD

Daly, A. (2019). Algorithmic oppression with Chinese characteristics: AI against Xinjiang’s Uyghurs. APC. Definition - Purchasing power standard | Insee. (2016). Retrieved 12 June 2020, from

https://www.insee.fr/en/metadonnees/definition/c1191

Deloitte. (2020). 2020 Technology Industry Outlook. Deloitte Center for Technology, Media and Telecommunications.

Deloitte China. (2019). Rising Innovation in China. Deloitte. Retrieved from

https://www2.deloitte.com/content/dam/Deloitte/cn/Documents/innovation/deloitte-cn-innovation-china-innovation-ecosystem-report-en-191101.pdf

Ding, J. (2018). Deciphering China’s AI Dream. Oxford: Centre for the Governance of AI, Future of Humanity Institute, University of Oxford. Retrieved from https://www.fhi.ox.ac.uk/wp-content/uploads/Deciphering_Chinas_AI-Dream.pdf

EU Member States sign up to cooperate on Artificial Intelligence - Shaping Europe’s digital future - European Commission. (2018). Retrieved 1 June 2020, from https://ec.europa.eu/digital-single-market/en/news/eu-member-states-sign-cooperate-artificial-intelligence

EPRS | European Parliamentary Research Service. (2019). Economic impacts of artificial intelligence (AI). Retrieved from

https://www.europarl.europa.eu/RegData/etudes/BRIE/2019/637967/EPRS_BRI(2019)637967_ EN.pdf

Eurostat Statistics explained. (2019). Gross domestix expenditury on R&D, 2006-2017 (%, relative to GDP) [Graph].

(31)

Executive Office of the President National Science and Technology Council Committee on Technology. (2016). PREPARING FOR THE FUTURE OF ARTIFICIAL INTELLIGENCE. NATIONAL SCIENCE AND TECHNOLOGY COUNCIL. Retrieved from

https://obamawhitehouse.archives.gov/sites/default/files/whitehouse_files/microsites/ostp/NS TC/preparing_for_the_future_of_ai.pdf

Furman, J., & Seamans, R. (2019). AI and the Economy. Innovation Policy and the Economy, 19(1), 161-191.

Gittleman, M., & Wolff, E. N. (1995). R&D activity and cross-country growth comparisons. Cambridge

Journal of economics, 19, 189-189.

Hagendorff, T. (2020). The ethics of Ai ethics: An evaluation of guidelines. Minds and Machines, 1-22. IC Insights. (2020). Top 10 Semiconductor R&D Spenders Increase Outlays 6% in 2017. Retrieved from

http://www.icinsights.com/news/bulletins/Top-10-Semiconductor-RD-Spenders-Increase-Outlays-6-In-2017

Independent High-Level Expert Group on Artificial Intelligence set up by the European Commission. (2019). A definition of Artificial Intelligence: main capabilities and scientific disciplines. High-Level Expert Group on Artificial Intelligence. Retrieved from https://ec.europa.eu/digital-single-market/en/news/definition-artificial-intelligence-main-capabilities-and-scientific-disciplines Jobin, A., Ienca, M., & Vayena, E. (2019). The global landscape of AI ethics guidelines. Nature Machine

Intelligence, 1(9), 389-399.

Knight, W. (2019). Why does Beijing suddenly care about AI ethics?. Retrieved 22 May 2020, from https://www.technologyreview.com/2019/05/31/135129/why-does-china-suddenly-care-about-ai-ethics-and-privacy/

Kok, J. N., Boers, E. J., Kosters, W. A., Van der Putten, P., & Poel, M. (2009). Artificial intelligence: definition, trends, techniques, and cases. Artificial intelligence, 1.

(32)

Li, L. (2018). China's manufacturing locus in 2025: With a comparison of “Made-in-China 2025” and “Industry 4.0”. Technological Forecasting and Social Change, 135, 66-74.

Living in the EU | European Union. (2020). Retrieved 13 June 2020, from https://europa.eu/european-union/about-eu/figures/living_en

Loucks, J., Hupfer, S., Jarvis, D., & Murphy, T. (2020). Future in the balance? How countries are pursuing

an AI advantage. Deloitte. Retrieved from

https://www2.deloitte.com/us/en/insights/focus/cognitive-technologies/ai-investment-by-country.html

Makridakis, S. (2017). The forthcoming Artificial Intelligence (AI) revolution: Its impact on society and firms. Futures, 90, 46-60. doi: 10.1016/j.futures.2017.03.006

McAfee, Andrew & Brynjolfsson, Erik. (2012). Big Data: The Management Revolution. Harvard business review. 90. 60-6, 68, 128

Mozur, P. (2019). One Month, 500,000 Face Scans: How China Is Using A.I. to Profile a Minority. Retrieved 13 June 2020, from https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html

Nunn, S. G., & Avella, J. T. (2015). Does Moral Leadership Conflict With Organizational Innovation?. Journal of Leadership Studies, 9(3), 85-87.

Park, S. (2017). The Fourth Industrial Revolution and implications for innovative cluster policies. AI &

SOCIETY, 33(3), 433-445. doi: 10.1007/s00146-017-0777-5

PricewaterhouseCoopers. (2016). Sizing the prize. Retrieved from

https://www.pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis-sizing-the-prize-report.pdf Qiang, X. (2019). The Road to Digital Unfreedom: President Xi's Surveillance State. Journal Of

(33)

Rankin, J. (2018). Artificial intelligence: €20bn investment call from EU commission. Retrieved 11 June 2020, from https://www.theguardian.com/technology/2018/apr/25/european-commission-ai-artificial-intelligence

School's facial recognition checks lead to fine. (2019). Retrieved 4 July 2020, from https://www.bbc.com/news/technology-49489154

Scholtens, B., & Dam, L. (2007). Cultural Values and International Differences in Business Ethics. Journal

Of Business Ethics, 75(3), 273-284. doi: 10.1007/s10551-006-9252-9

Schilling, M. (2017). Strategic management of technological innovation (5th ed., pp. 22-32). New York, NY: McGraw-Hill Education.

United Nations, Department of Economic and Social Affairs, Population Division (2019). World Population Prospects 2019, Volume II: Demographic Profiles

Wang, H. (2012). The conceptual basis of privacy standards in China and its implications for China's privacy law. Frontiers L. China, 7, 134.

Wladawsky-Berger, I. (2020). The Current State of AI Adoption. Retrieved 5 June 2020, from https://blogs.wsj.com/cio/2019/02/08/the-current-state-of-ai-adoption/

W. M. Cohen and D. A. Levinthal, “Absorptive Capacity: A New Perspective on Learning and Innovation,” Administrative Science Quarterly, March 1990, pp. 128–52.

World Bank national accounts data, and OECD National Accounts data files. (2020) Economic Policy & Debt: National accounts: US$ at current prices: Aggregate indicators. retrieved from

https://databank.worldbank.org/reports.aspx?source=2&type=metadata&series=NY.GDP.MKTP. CD

Zhu, J., Huang, T., Chen, W., & Gao, W. (2018). The future of artificial intelligence in China. Communications of the ACM, 61(11), 44-45.