Artificial Intelligence as your new colleague – How well do humans and machines collaborate?
A case study on the collaboration between humans and Artificial Intelligence in the Dutch banking sector
MSc in Business Administration, track: Digital Business
Name: Jeroen Lap
Student number: 11009136 Supervisor: Mw. Dr. R. Rotmans
EBEC approval number: 20210316070359 Word count: 14075
Statement of originality
This document is written by Student Jeroen Lap who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is 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.
This research's starting point is the literature gaps addressed by Seeber et al. in their 2020 research paper and Dahlin in her 2021 research paper
.Seeber et al. (2020) stress the importance of researching the various forms of collaboration between AI and humans in different fields.
According to Dahlin (2021), AI research should consider both the technological and sociological dimensions instead of focusing only on one dimension. To address both the research gaps of Seeber et al. (2020) and Dahlin (2021), this research looks into the following question: How does AI support team productivity in the Dutch banking sector? The research includes case studies at ING, ABN AMRO, and Rabobank to answer this research question.
The multiple case study consists of interviews with interviewees in different roles within the different Dutch commercial banks to gather information about the collaboration with AI and the team productivity. These different roles include end-users, managers, and developers. The research shows that AI supports the perceived team productivity by working together with employees. This collaboration contributes to the perceived team productivity if provided that the challenges related to collaboration are mitigated. The contribution can have multiped forms:
focused on quality improvement, improvement of the job experience, and the reduction of process time. The challenges companies need to consider are resistance and a low level of reliability. Factors benefiting the collaboration are training (non-technological factor) and the low usage threshold of an AI tool (technological factor).
Keywords: AI; Human-AI collaboration; Banking sector; Team productivity
Table of contents
1. Introduction ... 5
2. Literature review ... 10
2.1. Artificial Intelligence (AI) ... 10
2.2. Human-AI collaboration ... 11
2.3. Workplace innovation ... 13
2.4. Team productivity ... 14
2.5. AI in the banking sector ... 16
3. Research method ... 18
3.1. Research design ... 18
3.2. Research context... 20
3.2.1. ING ... 21
3.2.2. ABN AMRO ... 22
3.2.3. Rabobank ... 23
3.3. Data Analysis... 24
4. Case studies... 26
4.1. ING ... 26
4.2. ABN AMRO ... 30
4.3. Rabobank ... 35
4.4. Comparison ... 39
5. Conclusion & Discussion ... 44
5.1. Conclusion ... 44
5.2. Discussion ... 45
5.3. Limitations and Future Research ... 47
6. References ... 49
Appendix ... 55
A. Interview protocol ... 55
B. Interview questions written interview ... 56
C. Summary spoken interviews ... 58
D. Written interviews ... 61
List of Figures Figure 1 Thematic map... 25
List of Tables Table 1 Characteristics interviewees ... 19
Table 2 General information participating banks... 23
Table 3 Effects of the assistance of AI ... 40
Table 4 Challenges related to human-AI collaboration and mitigation measures... 42
The new way of working, where the machines execute and the people think. This collaboration between humans and machines may sound like a futuristic concept, but it is already part of the daily operation. The concept of Artificial Intelligence (AI) enables machines to be intelligent and (partly) autonomous (Makarius et al., 2020). To give an impression, 80% of the large companies are using a form of AI in 2020. AI is used in a lot of different industries, from surgical procedures in the health sector to fraud detection in the banking sector. The wide-scale integration of disruptive technological innovations (including AI) is considered the fourth industrial revolution (Makarius et al., 2020). The collaboration between humans and AI is one of the critical factors in determining the success of AI in the organization (Lichtentaler, 2018).
The interaction between humans and AI could boost the performance of the workforce or could lead to challenges due to a new complicated team dynamic (Makarius et al., 2020). The use of AI contributes to a more efficient, faster, and less erroneous work process in companies (Makarius et al., 2020). According to the 2018 research of Wilson and Daugherty, AI can support employees in several ways. A few examples, AI can amplify the analytic and decision- making abilities, interact with customers instead of the employees, and embody intelligence in machines that support humans (Wilson & Daugherty, 2018). These findings of Wilson and Daugherty are supported by the research of Johannessen (2018) and Loureiro et al. (2020). The (technological) advantages AI could bring are enormous and promising.
However, it is still the question of the potential of AI is really leveraged by organizations. Using AI is one thing, managing and overcoming the related challenges to AI is a different story. For instance, if employees are unwilling to collaborate with their machine counterparts, the maximum potential of AI will never be leveraged. Furthermore, even if humans and AI work together, it will not always result in the maximum contribution for the company due to the lack of trust humans could have in the ability of AI systems or the fear of replacement (Makarius et al., 2020). However, a well-managed human-AI collaboration could
provide the company with the opportunity to outperform other firms (Fountaine et al., 2019).
To identify the characteristic of successful and less successful collaborations between AI and Human a lot of additional research in various fields is required, according to Seeber et al.
The banking sector is one of the sectors where AI is widely adopted. The reason for this is primarily the willingness to work with technological innovations (Hassani et al., 2018). The Commercial banks leverage AI to deal with a large amount of data and automate a vast part of the decision-making (Du Croo De Jongh et al., 2018). According to the projections of Du Croo De Jongh et al., the commercial banks could benefit from an increase of 30% in revenue and a cost reduction of 25% in the 5 to 7 years after implementing AI on a large scale. An example of the application of AI in banking is Risk Management (Hassani et al., 2018). Risk management is one of the departments in banking where they use AI frequently, especially for detecting fraud. With the help of AI, employees are able to detect different fraudulent patterns.
Other AI applications in the banking sector are the deployment of AI to analyze consumer behavior for marketing purposes, the assistance of AI in customer support, and the use of AI in assessing potential loan clients (Hassani et al., 2018).
Besides all the advantages AI brings to banking, these advantages are accompanied by new challenges for the banking sector. One of the challenges is AI lacking the ability to mimic human intuition for decision-making (Huang et al. 2019). AI operates within a pre-programmed set of actions and is up to now not able to deal with intuitive thinking. In other words, Actions that require intuitive thinking needs to be performed by a human (Huang et al. 2019). Hence, at this moment, AI is not yet replacing the workforce; AI is collaborating with the workforce.
Although, there are employees who feel insecure about their position due to the increasing use of technological innovations such as AI (Liu & Zhan, 2020).
This research's starting point is the literature gaps addressed by Seeber et al. in their 2020 research paper and Dahlin in her 2021 research paper. Dahlin criticizes the lack of research
projects involving more than one perspective. According to Dahlin (2021), AI research should take both the technological and the sociological dimensions into account instead of focusing only on one dimension. Seeber et al. (2020) stress the importance of researching the various forms of collaboration between AI and humans in different fields. The researchers consider their own research as the beginning of the chain of new research on human-AI collaboration (2020). Different researchers and research papers support the call for more research on collaboration with AI, including Loureiro et al.’s 2020 research paper. Loureiro et al. (2020) are known for mapping and keeping track of the different AI research projects. The researchers suggest the impact of AI on the workforce as a future research topic. To address both the research gaps of Seeber et al. (2020) and Dahlin (2021), this research looks into the following question:
How does AI support team productivity in the Dutch banking sector?
To answer the research question, this research includes a multiple case study within the banking sector in the Netherlands. The multiple case study consists of interviews with interviewees in different roles within the different Dutch commercial banks to gather information about the collaboration with AI and the team productivity. These different roles include end-users, managers, and developers. In order to provide a final answer to the research question, this research will answer a few sub-questions first.
The strengths of AI, the different ways AI could benefit to organizations is the first concept this research would look into. According to the research paper of Miller (2019), the development of AI's abilities is astonishing. Although Miller (2019) also recognizes the current limits of AI and the necessity of humans still being involved in performing certain tasks. The question arises on how the current contribution of AI looks like. Based on the existing literature and information gathered in the field, this research will answer one of the following sub- questions: In which ways can AI assist employees in their work?
The second concept this research needs to gather information on is the weaknesses of AI. In other words, which social and technological challenges could occur when working together with AI counterparts or while implementing AI in the organization. An example of a challenge named in the academic literature is the resistance of employees to work with new technology (Ali et al., 2016). The sub-question this research would like to answer is: What challenges occur when working together with AI and how to overcome these challenges? It is essential to know what challenges could occur because companies then have to prevent the challenges from occurring or resolve them immediately.
The main objective of this research project is to provide more clarity on to how AI can enable teams to be productive and how AI can support the employees in their work.
Furthermore, this research also provides insights into the challenges that could occur when working together with AI. Simultaneously, this research project on the productivity of human- AI teams in the banking sector also addresses the need for research on collaboration in specific fields (Seeber et al., 2020) and the request for more research covering both the technological and social aspects of human-AI collaboration (Dahlin, 2021). This master thesis is an addition to current AI literature about Human-AI collaboration. By adding the observations of three case studies in the banking sector, this master thesis provides additional examples of factors improving the collaboration as well as factors deteriorating the collaboration between humans and AI.
Another intention of this research is to advise organizations on leading teams consisting of both humans and machines. To show organizations how the Dutch bank sector deals with team-related challenges and tries to maximize the productivity of these teams. The way some of the Dutch banks applying AI might be used as a best practice for other companies leveraging AI. This report's outcomes could also be used as an inspiration for companies, which are not working with AI yet or failing to increase the productivity of their teams.
The thesis will be structured as follows: In the second chapter, the literature will be reviewed related to AI, workplace innovation, team productivity, Human-AI collaboration, and AI in banking. In the third chapter, the research method and units of analysis will be discussed.
The case studies will be presented in the fourth chapter. Finally, the implications of these case studies will be the subject of the fifth chapter, the conclusion, and the discussion.
2. Literature review
The request of Seeber et al. (2020), the call from Dahlin (2021), and the suggestion made by Loureiro et al. (2020) combined lead to the research topic of this master thesis: The team productivity of a team consisting of humans and AI in the banking sector. Before discussing the research setup, five crucial concepts related to the research topic will be addressed. It concerns the following concepts: Artificial Intelligence, human-AI collaboration, workplace innovation, team productivity, and AI in the banking sector.
2.1. Artificial Intelligence (AI)
The central technological concept in this research is Artificial Intelligence. There are a lot of different definitions of Artificial Intelligence. This research is based on the definition by Gallagher. In his book, Gallagher described Artificial as follows: “Computer software that can mimic or improve upon functions that would otherwise require human intelligence” (2019). The definition covers a variety of branches and categories. One of the most known branches of AI is Machine learning. Machine learning includes all the software which is able to learn without interventions by the programmer (Gallagher, 2019). In particular, Data scientists leverage Machine Learning in their models in order to analyze tons of data. An important distinction is the fact that models are supervised or unsupervised. A Supervised model is trained with a training set of data to improve the outcome. An unsupervised model is detecting patterns without having any training input (Gallagher, 2019). Examples of AI applications are Natural Language Processing (NLP), Speech Recognition, and Pattern Recognition. In the research, the accent is not on one application of AI in particular but on the vast array of AI applications in the Dutch banking sector. In the last section of this chapter, the different applications of AI in the banking sector are being explained.
Nowadays, the potential of leveraging AI tools is far-reaching. The common observations on the potential of AI are based on the broad application of AI technology and the
high expectations of the future developments of AI (Miller, 2019; Wilson & Daugherty, 2018).
Barro and Davenport (2019) acknowledge the potential of AI, but in their opinion, the potential will only be leveraged as the company uses AI to innovate. Furthermore, attracting employees with the right technical skills to accomplish innovation is also an important condition.
The question is not whether AI will take over tasks from humans, but the question is which tasks (Miller, 2019; Wilson & Daugherty, 2018)? For example, different researchers express their concerns about the abilities of AI to perform tasks involving emotions (Lichtentaler, 2018;
Huang et al., 2019). Other shared concerns about the future application of AI relate to ethical questions, the safeguarding of sensitive information, and the mitigation of bias (Davenport, 2019; Stahl & Wright, 2018).
As illustrated by the 2020 paper of Makarius et al., many companies are already exploiting the potential of AI. Although the maximum performance AI can achieve is many times higher than humans ever could, leveraging AI is still not without challenges. One of these challenges is the communication between AI and humans (Kaplan & Haenlein, 2020). From the side of AI, the communication problem occurs in the translation of the demands of humans. The AI application or model is not able to understand the wishes of the user (Kaplan & Haenlein, 2020).
From the side of humans, the communication problem is understanding the process the AI application executes. That is why the user cannot validate the process or explain the process steps to others, for instance, to clients (Kaplan & Haenlein, 2020). Additionally, there is still the other major challenge of AI being biased. This is further discussed in the section about the application of AI in banking.
2.2. Human-AI collaboration
The main subject of this thesis is the collaboration between humans and Artificial Intelligence.
As mentioned in the introduction, the collaboration between humans and AI is already part of the daily operation of many large companies (Makarius et al., 2020). Furthermore, there is a
shift in the human/digital configuration. Office work is increasingly performed by digital systems, like Artificial Intelligence (Baptista et al., 2020). The ability of AI to assist humans is also proven by different researchers, among others Wilson and Daugherty, Lichtentaler, and Fountaine et al. Wilson and Daugherty describe the three various forms of AI supporting humans: amplify the analytic and decision-making abilities, AI can interact with customers, and AI can embody intelligence in machines that support humans (2018)
However, the interplay between humans and AI not always result in reinforcing one another, especially if emotional intelligence is required (Lichtentaler, 2018; Huang et al., 2019).
The lack of emotional intelligence of AI is one of the technological challenges related to the collaboration between AI and Humans. The lack of emotional intelligence could hurt the reliability of an AI tool (Huang et al., 2019).
Next to this technological challenge, there are a lot of non-technological challenges complicating the relationship between humans and AI, for example, the fear of employees being replaced by AI or the lack of trust in AI (Fountain et al., 2019). The research of Baptista et al.
AI supports the conclusions of Fountaine et al. about the non-technological challenges (2020).
Although, in the first place, the study of Baptista et al. also supports the beneficial effect of the human-AI collaboration. Another non-technological challenge is the lack of training of the employees who are going to work with the AI tool (Ali et al., 2016). The lack of training could lead to resistance among the employees.
Earlier research on the collaboration between humans and Artificial Intelligence of Amershi et al. (2019) provide some insights on how to improve the collaboration. Before Humans and AI starting to work together, it is vital to convince the end-user of the numerous possibilities of AI. Furthermore, it is important to show AI's performance level on the different applications (Amershi et al., 2019). When AI and humans work together, AI needs to provide an understandable explanation to humans about executed actions (Amershi et al., 2019).
Besides giving explanations, AI should perform actions matching the social norms when taking
over human’s tasks. This includes avoiding the execution of actions based on any bias (Amershi et al., 2019). Amershi et al. (2019) conclude their paper by stressing both the importance of continuously giving feedback about the collaboration and the importance of having any form of oversight in place.
In the current academic literature, there is much known about humans' upsides and downsides and AI working together. On the other hand, there is less known about managing both the collaborations' strengths and weaknesses. Although, according to the paper of Kolbjørnsrud et al. (2017), managing is essential in order to create ideal conditions for the collaborations to be productive for the company. The research of Kolbjørnsrud et al. (2017) describes three steps for the manager to turn AI into a success. The First step is to start experimenting with AI. In other words, exploring the different applications of AI. The second step is to track the performance of AI, especially the performance in privacy-sensitive processes. The final step is to compose a training plan for the potential end-users. These steps together should lead to more support base for AI and, therefore, to better Human-AI collaboration (Kolbjørnsrud et al., 2017).
Finally, leveraging the human-AI collaboration is also beneficial because of AI’s supportive role in workplace innovation. The reason for a company to invest in workplace innovation is that it improves the company performance and the quality of the jobs simultaneously (Oeij et al., 2012). So, both the company and the employee benefit from AI’s role as supportive technology. As this research focuses on the experiences of employees working with AI, the quality of the jobs is touched upon as well. In the next section, the concept of workplace innovation is further explained.
2.3. Workplace innovation
The third concept of interest is workplace innovation. According to Oeij et al. (2012), the definition of workplace innovation is “The implementation of new and combined interventions
in the field of working organization, HRM and supportive technologies. Workplace innovation is considered to be complementary to technological innovation”. The link with the research topic is the so-called supportive technologies. AI could fulfill the role of being the new supporting technology (Wilson & Daugherty, 2018). However, if those supporting technologies are not well managed, it could lead to less performance. The objective of workplace innovation is to both increase the performance of the company and the quality of jobs (Oeij et al., 2012).
In the research of Pot (2011), the performance of companies improved after implementing different types of work innovations. The innovations also contributed to more appreciation of the quality of the job. Totterdill and Exton (2020) emphasize the importance of the workforce embracing supportive technologies instead of feeling the fear of replacement. The leverage of supporting technologies could create new opportunities and new jobs. A side effect of implementing workplace innovations is the change in required competencies of employees who need to adapt to the changing workplace. If employees are not able to adapt to the innovations, they will be replaced by employees matching the required competencies (Johannessen, 2018).
To prevent employees from failing to adjust to the new workplace innovations, employees should receive guidance in adapting to the new required competencies (Totterdill and Exton, 2020).
2.4. Team productivity
Team productivity is another important concept for this research project. The team productivity shows us the performance of a particular team and enables researchers to compare different team compositions. The reason why this research focuses on team productivity instead of productivity is to respond to the call of Dahlin (2021). By including the team component, this research focuses on the social dimension and not solely on the technological dimension.
Working as a team is, under normal circumstances, more productive than working on your own (Mendelsohn, 1998). For a team to function, three conditions need to be fulfilled. It
concerns the following three conditions: the team has a team leader, team members are willing to be led, and discipline is present (Mendelsohn, 1998). Besides the general research of Mendelsohn (1998), there is a lot of research on different influencing factors of team productivity. For example, personal characteristics, work environment, and the type of work.
The influence of technology on team productivity is also researched. According to the research of Tohidi and Tarokh (2006), investments in information technology (IT) are necessary to increase the maximum team productivity cap. Research on the contribution of advanced technologies such as AI to productivity in the service sector shows a positive contribution (Rust
& Huang, 2012). In this research project, the focus is on the influence of AI on perceived team productivity.
Measuring team performance in the manufacturing industry is relatively straightforward. Researchers and managers only have to look at the physical output. In contrast to the service industry, where measuring the output of knowledge workers is more difficult and complex (Ramírez & Nembhard, 2004). At the end of the previous century, the researcher Peter Drucker considered determining the productivity of the knowledge workers as the “biggest”
challenge for business researchers in the 21st century (1999). In the research of Ramírez and Nembhard, multiple solutions are discussed for measuring the productivity of knowledge workers, including both quantitative as qualitative measures. Conducting interviews is a qualitative method often used to determine the (team) productivity.
However, different researchers are still looking for the most suitable way to determine productivity for knowledge workers. For example, Moussa et al. (2016) developed a questionnaire to assess productivity. The questionnaire is based on different factors influencing the productivity of workers, think of well-being, and knowledge sharing (Moussa et al., 2016).
In the recent article of Sookdeo (2020), another perspective to determine the productivity of knowledge workers is offered. Sookdeo defines productivity as the success of the company. In his opinion, investing in education led to more skilled workers, who contribute more to its
success and, therefore, will be considered more productive. As there is no consensus about measuring productivity, this research will base the (team) productivity on the interviewees' experiences.
2.5. AI in the banking sector
The last concept is the application of Artificial Intelligence in banking. In the introduction of this proposal, the research of Du Croo De Jongh et al. (2018). described the widespread adoption of AI in the banking sector. Moreover, the researchers emphasized the potential (financial) benefits for commercial banks, which implement AI in their companies.
Next to the potential benefits of AI in the banking sector, the possible application within the banking sector was touched upon briefly in the introduction. Hassani et al. described the strengths of AI in detecting fraudulent patterns within the field of risk management. (2018).
The application of AI described by Hassani et al. is a form of machine learning, a branch of the field of AI (Selvaraj, 2019). Another application of AI within the banking sector is the virtual assistant, a tool deployed to assist people with their support questions. The virtual assistant falls in the domain of natural language processing (NLP), a branch within the AI focused on translating speech and text (Selvaraj, 2019). The wide variety of applications of AI contributing to the bank working smarter. For instance, machine learning could be leveraged to automate different processes, from approving commercial loans to advising the customer about trends in stock trade. Furthermore, NLP could be leveraged be to assist the support teams of commercial banks (Akhilesh & Möller, 2019)
Though, the application of AI in banking comes along with some challenges. In the article of Howard, she addresses multiple of the big challenges related to the application of AI.
Banking systems using AI technology could become biased in making decisions, might lack the preferred level of transparency, or possibly violate the privacy of clients (Howard, 2019).
Therefore, Howard advocates for regulation for companies using Artificial Intelligence. She
stresses the many advantages of AI but at the same time asks for awareness of the impact poor regulated AI could have (2019).
The forms of collaboration between humans and Artificial Intelligence in banking differ per process (see the applications mentioned above). In some cases, AI is fully taking over the tasks of the human counterpart, for instance, a chatbot on the bank's website (Deutsche Bank Research, 2019). In other cases, humans and AI work on the same tasks together. For example, AI detects a fraudulent pattern in a dataset, and humans judge the correctness of the fraud indication (Deutsche Bank Research, 2019). In the report of the Deutsche Bank Research (2019), the researchers expect AI to take over more tasks in the future once AI tools are more developed.
3. Research method
In this chapter, the research design is discussed, the research environment is explained, and the units of analysis are introduced. Further, the process of data analysis is set out.
3.1. Research design
This research project on the collaboration between AI and Humans follows a multiple case study as a research strategy. A multiple case study is a suitable way to explore how banks leverage the collaboration between AI and humans to support team productivity. According to Yin (2009), case studies are suitable to help to explain a complex social phenomenon. In this research, leverage the collaboration between AI and humans. Research question starting with why or how are a good fit for case studies (Yin, 2009). In this research, the focus is on how AI supports team productivity. The Dutch banks are the units of analysis in this multiple case study.
Due to the time constraint of the thesis and the COVID-19 restrictions, this research focuses on three Dutch banks. Observing the three Dutch banks should be enough to give an impression of the Dutch banking sector’s human-AI collaboration. One team that uses AI at each bank is the unit of analysis, a so-called holistic case study design. The multiple case study reflects the state of the collaboration at different banks at a certain point in time and is, therefore, a cross- sectional case study. This multiple case study approach is inductive because testing existing theories is not part of this research. Furthermore, constructing possible theories/statements comes after the observations.
This research project's required data is gathered by interviewing bank employees working with AI Tools and managers who lead teams working with AI( A translated version of the interview protocol is included in appendix A). The reason for including the managers is to show different perspectives and the opportunity to see if these experiences match with their team members. The initial plan was to interview per bank two employees and one manager (nine interviewees in total). Unfortunately, due to the Covid-19 pandemic, it was more difficult
for banks to make people available. Therefore, the total number of interviewees is seven. Five of the interviewees are employees, and two are managers. The characteristics of the interviewees are presented in table 1. The interviews were conducted in Dutch to allow the interviewees to answer in their native language. All the interviewees answered written questions, a so-called written interview. The written interviews consist of fifteen open questions (A translated version of the written questions is included in appendix B). At each bank, there was also a spoken interview with one of the interviewees. The spoken interviews last at least around thirty minutes to assess the experienced productivity within the bank (A summary of the spoken interviews is included in appendix C). This interview will consist of questions about the experience of working with AI, the advantages of collaboration, and the possible challenges in working or implementing AI. The answers to the interview questions are used to help us understand in which ways AI can support team productivity. The interview will also be used as a subjective determinant for the productivity of human-AI teams. As input for the case studies, the spoken interviews outweigh the written interviews. The reason for this is that a spoken interview gives the opportunity to ask follow-up questions and to experience the emotions of the interviewee. The possibility to ask follow-up based on the written interviews was limited due to the availability of the interviewee. In the case of follow-up questions based on the written interviews, these questions often are addressed in the spoken interviews.
Table 1 Characteristics interviewees
Interviewee Bank Profession Years at
the bank Relation to AI
[Employee 1] ING Private banker 27 Collaborating with an AI tool [Employee 2] ING Private banker 18 Collaborating with an AI tool [Employee 3] ING Private banker 18 Collaborating with an AI tool [Employee 4] ABN AMRO AI developer 2 Developing AI tools, Monitoring usage [Employee 5] ABN AMRO Data scientist 3 Developing AI tool, Monitoring usage
[Manager 1] ABN AMRO Manager
Innovation 38 Directing AI development team, Communication with stakeholders AI [Manager 2] Rabobank Business
Data 14 Directing AI development team, Communication with stakeholders AI
Conducting interviews is the only data collection method used in this research, so this research follows a mono qualitative data collection design (Saunders, 2015). The interviewees are selected according to a non-probability sample strategy, more specifically, a self-selection strategy. The interviewees were suggested through acquaintances from the (business) network.
In case the profile matched, the interviewees were invited to take part in the research. To augment the confidence of the data collected in this research, triangulation could be a solution.
Adding a second researcher or executing quantitative analysis as well are forms of triangulation that could be applied. Only due to time and capacity constraints triangulation is not possible for this research.
3.2. Research context
The reason for specifically picking the banking sector as a research environment is the great willingness to work with new technological innovations, among others AI (Hassani et al., 2018). In the research of Kauer et al. (2020), different reasons for (frequently) using AI in the banking sector are named. For instance, to cope with the enormous amount of data, detect fraud sooner, offer more self-service, deliver more efficient processes, and augment the productivity of employees. Another reason to focus on the banking sector is the trend of banks replacing the current infrastructure with digital alternatives. For example, the closure of local banks is a consequence of increasing banks' app possibilities (Banken.nl, 2021). An interesting question could be: What role does AI play in this trend? The focus is on the Dutch banking sector because of its feasibility. Time and capacity constraints forcing this research to be limited to the Dutch banking sector. Focusing solely on the Dutch banking sector has one advantage, namely, the interviewees are able to share their stories in their native language.
The objective of this research project is to explore the ways Artificial Intelligence supports team productivity in the Dutch banking sector. To find out how the Dutch banks leverage the potential of AI and deal with the associated challenges, a multiple case study is
performed. The three Dutch banks selected for this multiple case study are ING, ABN AMRO, and Rabobank. These three banks have an aggregated market share of 82% of the Dutch banking market (Banken.nl, 2020). Below ING, ABN AMRO, and Rabobank are further introduced. In particular, the views of these banks on the importance of technological initiatives are discussed. An overview of the general information of the participating banks is presented in table 2 (on page 23).
ING is the largest bank of the Dutch banking sector (39% market share) (Banken.nl, 2020). The annual report of 2020 shows a net profit of almost 2.5 billion euro (ING Group, 2021). An increase of over 700 million compared to the year 2019. The roots of the ING bank in the Netherlands can be traced back to 1881 when the Dutch government established the Rijkspostspaarbank (ING Group, n.d.). The Rijkspostspaarbank enabled the Dutch society to save money in a safe way. Over the years, ING expanded to more domains, like payments, mortgages, small and medium enterprises (SME’s), and wholesale banking (ING Group, n.d.).
Nowadays, ING has more than 56,000 employees and serving around 39 million clients in 40 different countries.
In the 2020 annual report of ING, the importance of going digital and data is emphasized multiple times. The CEO of ING, Steven van Rijswijk, declares digitalization and data as the two key pillars of the strategy to become future-proof (ING Group, 2021, p.13). Data is an important part of fraud detection and prevention. Digitalization is the main solution for improving the experience of customers (ING Group, 2021). In the process of becoming future- proof, AI plays an important role as well, among others at detecting fraudulent patterns. To leverage the potential AI offers ING partners up with third parties, for instance, the tech company Tradeteq and the University of Technology Delft (ING Group, 2021).
3.2.2. ABN AMRO
ABN AMRO is the third bank of the Netherlands in terms of market share (17%) (Banken.nl, 2020). In 2020 the bank reported a loss of 45 million (ABN AMRO Bank N.V., 2021a). The drop in net income is more than 2 billion compared with the reporting year 2019 (profit of 2 billion). The loss is predominantly caused by the COVID-19 pandemic, which impacted both the results of the bank and its clients (ABN AMRO Bank N.V., 2021a). ABN AMRO is established in 1991 after a merger of Algemene Bank Nederland (ABN) and Amsterdam- Rotterdam Bank (AMRO) (ABN AMRO Bank N.V., n.d.). The Dutch government owns half of the shares of ABN AMRO. The government owning shares stems from the nationalization of ABN AMRO in 2008, which was necessary to save ABN AMRO from the effects of the financial crisis (ABN AMRO Bank N.V., n.d.). Currently, ABN AMRO has 19,000 full-time employees worldwide working in the different domains of the bank. The different domains within ABN AMRO are Private Banking, Commercial Banking, Retail Banking, and Corporate
& Institutional Banking (ABN AMRO Bank N.V., 2021a).
The (future) focus of ABN AMRO is on leveraging more technology to serve the customer optimally and personally in this digital era, for example, offering the client the possibility to have a video call with an advisor (ABN AMRO Bank N.V., 2021b). The bank aims to digitalize more processes in the foreseeable future in order to provide clients with more services through digital channels. ABN ARMO also uses digital solutions to make internal processes more efficient. The bank stresses in the annual report the need for more digitally skilled employees (ABN AMRO Bank N.V., 2021b). Having a digitally skilled workforce is a necessary strategical asset for ABN AMRO because the bank expects big tech firms may enter the banking market in the future (ABN AMRO Bank N.V., 2021b).
Rabobank is the second-largest bank in the Netherlands. Rabobank has a market share of 26%
(Banken.nl, 2020). Over the year 2020, Rabobank has a financial result of 1 billion profit (Rabobank, 2021). Despite the positive result, COVID-19 caused the profit of Rabobank to halve. The history of the Rabobank starts in 1895, when multiple small banks form two large corporations, namely Coöperatieve Centrale Raiffeisen-Bank and Coöperatieve Centrale Boeren-leenbank (Rabobank, n.d.). These corporations are closely intertwined with the agricultural sector. These corporations merged in 1972 and named itself the Rabobank (Rabobank, n.d.). Rabobank is present in around 40 countries and has 10 million customers.
Two-thirds of the customers of Rabobank are living in the Netherlands (Rabobank, n.d.). In the 21st century, Rabobank is still partly focused on the agricultural sector.
Regarding the application of technological innovations, Rabobank has the intention to invest more in these technological innovations, especially to detect fraudulent transactions (Rabobank, 2021). Rabobank is improving the customer experience by upgrading digital channels, for example, by extending the functionalities of the mobile app. The bank is planning to continue upgrading its digital channels in the coming years. In the annual report, Rabobank mentions the risk associated with leveraging more technology. Therefore, the bank is investing more money in monitoring the behavior of the applications (Rabobank, 2021).
Table 2 General information participating banks1
Name ABN AMRO ING Rabobank
Date of establishment 1991 1881 1972
Headquarters Amsterdam Amsterdam Utrecht
Net profit Loss of 45 million 2.5 billion profit 1 billion profit
Market share 17% 39% 26%
Number of employees ± 19,000 ± 56,000 ± 43,000
1 based on the sources: (ABN AMRO Bank N.V., 2021a), (Banken.nl, 2020), (ING Group, 2021) &
3.3. Data Analysis
After gathering the data of the written and spoken interviews, the spoken interviews are transcribed. Subsequently, the transcribed and written interviews are imported into the analysis program NVivo in order to code the data. The data analysis is based on the thematic data analysis method of Braun and Clarke (2006). After the initial step of getting familiar with the data, the second step is to assign initial codes to important features in the transcriptions and written interviews. This set of initial codes primarily consists of codes about the value of the deployment/collaboration AI and the challenges related to AI. For the example, the following quote is labeled as increase in quality: “het advies is absoluut van hoger niveau, hè, omdat je veel beter inzichtelijk kan maken hoe de situatie is”. The third step of the method of Braun and Clarke (2006) is to detect the different themes within the set of codes. As mentioned earlier, the code set consists primarily of codes about factors favoring the collaboration between humans and AI and codes about factors inhibiting the collaboration between humans and AI. Therefore, benefiting factors on the collaboration and inhibiting factors on the collaboration are the two themes within the set of codes. The code “increase in quality” is related to the benefiting factors theme. The overarching theme is the contribution of the human-AI collaboration to perceived team productivity. In the next steps of the analysis method, it is important to review the themes and draw a thematic map (Braun & Clark, 2006). The thematic map is presented in figure 1 on the next page. In the review step, there is a check on the fit between the themes and the entire collection of data. Additionally, the theory on the collaboration between humans and AI is compared to the set themes. The themes match with the existing theory. In the academic literature on collaboration, both the beneficial aspects as related challenges are discussed. These codes and themes are the basis of the case studies and the comparison in the fourth chapter.
Figure 1 Thematic map
The perceived contribution of human-AI
collaboration to team productivity
Reduction in process time
Increase in quality
Increase in job experience
Lack of training
4. Case studies
In this chapter, the case studies on the three largest Dutch banks are discussed. The first case study is about ING’s application of AI in the private banking department. In the second case study, the focus is on the application of AI within ABN AMRO. A current project in the customer contact center is used as an example. In the third case, the deployment of different ML models throughout Rabobank is the main topic. Each case study concludes with a future outlook on the deployment/developments of AI (within the bank). Finally, in the fourth paragraph, the similarities and differences between the case studies are explained. In the case studies, the names of the interviewees are anonymized in order to comply with the (privacy) regulations of both the participating banks and the University of Amsterdam.
In this first case study, the experiences of the private banking department of ING with AI are analyzed. The private bankers of ING are working with an in-house developed AI tool called Forward Planning. The primary purpose of Forward Planning is to provide insights into (potential) return on investments. This tool is leveraged by ING to support the private bankers of ING in convincing clients of the (future) benefits of private banking and to keep track of the current financial performance. To learn about the experiences of Forward Planning, three private bankers are willing to share their experiences. To refer to statements/experiences of these bankers, the references [Employee 1], [Employee 2], and [Employee 3] are used.
The application of Forward Planning in the meetings with (potential) clients has a strong additional value, an observation shared by all the three private bankers. [Employee 1] considers Forward Planning a competitive advantage over other banks (all answers of the written interview are present in appendix D). [Employee 2] experiences the following benefit of collaborating with Forward Planning:
“Het geeft zowel de klant als de medewerker duidelijkheid en comfort welke beslissingen er moeten worden genomen.”.
In this Dutch quote, [Employee 2] tells about the experience of more comfort for both the customer and the private banker when making financial decisions together. In other words, the quality of the meetings improves. The benefit of more clarity for all the participants is the main contribution of the AI tool, according to [Employee 3]. [Employee 3] underlines the additional value of the tool for clients, advisors, and the bank with the following quote:
“Ik zie het als toegevoegde waarde voor klant, adviseur en bank.”.
In the spoken interview with [Employee 1], she expressed her enthusiasm by giving the following response:
“Ja, ik was meteen enthousiast” (personal communication, May 5, 2021).
This is a Dutch answer for expressing your immediate enthusiasm. The response of [Employee 1] shows that working together with AI also benefits employees' job experience.
Another beneficial factor to the contribution of the collaboration between humans and AI is the training the participants received. In preparation for working with Forward Planning, the private bankers participated in two introduction days. The purpose of the training is to safeguard the quality of the provided services and guarantee a smooth collaboration.
[Employee 3] tells in the interview about his training experience:
“Wij hebben twee opleidingsdagen gehad en daarna een toets met een klantcasus, waarbij de financieel planner een klant acteerde en de leidinggevende beoordeelde.”.
In the introduction days, the future users of Forward planning were trained on different functionalities and had the opportunity to practice in a test environment. This training cycle concluded with an exam, in which the participants had to show their ability to work with Forward planning. The importance of the training days is stressed by [Employee 1]. In the interview, [Employee 1] tells about the difficulty of adapting for the employees who lack digital skills. Something she experiences herself:
“Als je niet heel digitaal ingesteld bent, is het lastig om onder de knie te krijgen.”
She lacked the required digital skills and failed twice before passing the introduction exam (personal communication, May 5, 2021).
Next to the beneficial factors, there are also factors inhibiting/challenging the collaboration between humans and AI and, therefore, hurting the contribution to the team productivity. For instance, the inhibitory factor resistance. Some ING employees are not happy to work with AI. According to [Employee 1], this unhappiness is caused by the fear of replacement. In the opinion of [Employee 3], some employees resist because the AI tool interferes with the commercial process.
On the other hand, [Employee 2] is not aware of any form of resistance among the private bankers. Another group showing signs of resistance is the group of older employees. This group was initially not convinced of the necessity of the new AI tool. This observation stems from the spoken interview with [Employee 1]. She mentioned the following;
“Bankiers van de oude stempel die zoiets hebben van: “Joh, ik heb dit helemaal niet nodig”.
This Dutch quote implies the older employees considering the new technology as unnecessary.
However, in the experience of [Employee 1], this opinion disappears after working with the tool and experiencing the benefits. She supports this with the following dutch quote:
“als je het effect ervaart, dan kun je d'r niet op tegen zijn.”.
Besides the issues about resistance, problems with the reliability of the AI tool could also have an inhibitory effect on the collaboration. The three private bankers are also sharing their views about the reliability of the AI tool. All three private bankers have a positive experience regarding Forward Planning. [Employee 1] qualifies the reliability as good, but she does share a technical concern about the extensive network load of the tool. [Employee 2]
illustrates the level of reliability with a comparison between the actual returns and the predicted returns: “Ervaring is prima en de motor waarop het programma draait (Ortec) is afgelopen tijd defensiever geweest als de uiteindelijke rendementen van vermogensbeheer.”.
Compared to the actual results, Forward planning predicted a more defensive outcome. The more defensive predicting of Forward planning contributes to the reliability. [Employee 3]
stresses again the positive contribution of forward planning to the quality of the meetings with clients.
About the central aspect in this research project, team productivity, the private bankers do not have a unilateral opinion. According to [Employee 1], the experienced team productivity after the introduction of the AI tool does not concern an increase in the number of meetings but an increase in the quality of the meetings. For example, more comfort in making financial decisions. In the spoken interview with [Employee 1], she addresses the higher quality meetings with the following quote :
“het advies is absoluut van hoger niveau, hè, omdat je veel beter inzichtelijk kan maken hoe de situatie is”.
In this quote, [Employee 1] speaks about the higher quality meetings because of the ability to provide better insights into the financial performance of the client. In the opinion of [Employee 2], the clear organized design of Forward Planning enables more employees to meet with clients to offer financial advice. An additional benefit of Forward Planning [Employee 2]
mentioned in his interview is the better recording of information than before the introduction of the AI tool. In contrast, [Employees 3] does not agree with more experienced team productivity. In his experience, the tool is not explicitly contributing to perceived team productivity:
“Niet veel, je bent een voorstander van het programma of je vindt het een storende factor.”.
This Dutch statement of [Employees 3] means that you are either a supporter of the program or you just find it a disruptive factor. At the same time, [Employee 3] acknowledges the additional benefits for the client, like more certainty and rest about the financial performance.
The private bankers welcome a future in which more digital techniques are used at the bank. In the interview with [Employee 1], she considered the trend of replacing physical
processes with digital alternatives as ING’s top priority. An example of this replacement is the closure of the majority of the local banks due to an extension of the functionalities in ING’s app. Regarding the future development of AI tools, [Employee 1] expects that clients are able to use future AI tools themselves without the help of an employee from the bank. This statement is derived from the following Dutch quote:
“Ik vermoed dat we in de toekomst ook vaker tools zullen worden ontwikkeld die de klant wel zelf kan gebruiken.”.
[Employee 2] finds a future with the application of more digital techniques a positive development, provided that the human dimension is always safeguarded. The additional value of applying AI on a greater scale is underlined by [Employee 3]. In the interview with [Employee 3], he explains the additional value of Artificial Intelligence by highlighting one of the strengths. According to [Employee 3], the strength of the AI tools is being not emotionally involved in a decision, while the clients most of the time are emotionally involved.
4.2. ABN AMRO
In the second case study, the focus is on the application of AI within ABN AMRO. To gather information about the application AI within the bank, ABN AMRO was willing to make two members and the managers of the AI development team available for interviews. Next to developing AI tools, this team guides and monitor the end-users of AI closely. For instance, the agents of the customer contact center. The team members share their vision on the value of AI and the related challenges. The manager of the AI development department illustrates the application of AI with an example of a current project. This project is about an AI tool deployed in the customer contact center to assist the support agents. The AI tool supports the agents by providing multiple solutions/answers, which the agents can use as a response to the customer.
The purpose of this AI tool is to shorten the response time of agents. To refer to the statements
and experiences of these interviewees, the references [Manager 1], [Employee 4], and [Employee 5] are used.
Based on their own experiences and the shared experiences of the agents of the customer contact center, all the interviewees consider the introduction of AI in the bank as a big step forward. [Employee 4] and [Employee 5] both think that AI’s most significant benefit is cost reduction due to the automation of several processes (all answers of the written interview are present in appendix D). In the specific example of the customer contact center, the main benefit is the reduction in response time as told by [Manager 1]:
“Nou dan hebben ze toch al gauw 5/10 seconden tijd bespaard en dat dat loopt nu heel erg op.”
In this quote, [Manager 1] shares the measured results of a 5/10 seconds reduction in response time, and this reduction is still increasing. [Manager 1] also shares the following quote of a customer support agent who is experiencing unavailability of the AI tool:
“My timesaver doesn't work, can you please fix it” (personal communication, April 22, 2021).
[Manager 1] used this quote to illustrate how important the AI tool already is, just a few months after the implementation. According to [Manager 1], another potential benefit of applying AI in customer support is the ability to answer some questions autonomously. In the spoken interview with [Manager 1], he expects the AI tool to be able to answer some questions autonomously at the end of 2021.
Regarding the influence of training, [Employee 4] indicates that he has not received any training at all before starting to work with AI. Instead of being trained, [Employee 4] had to find out himself. In contrast, [Employee 5] received multiple training to develop AI tools and, more specifically, ML models. The support agents were trained and monitored in the pilot phase of the implementation. In the interview, [Manager 1] tells about the low usage threshold to use the AI tool:
“Die gebruiksdrempel is heel laag … alleen de drempel om het op een goede manier in je scherm te krijgen. Die is best wel hoog. In ons geval moest die interface ook handmatig door developers worden aangepast”.
In this quote, [Manager 1] confirms the low usage threshold and explains how developers resolved the issue regarding the display of the answer options. In other words, developers have an important role in keeping the AI tool's threshold low. Next to the agents being trained to work with the AI tool, the tool is also being trained by the agents as proudly told by [Manager 1]:
“Wij gaan daar nog een stap verder in, wij gaan de mensen ook de downvote mogelijkheid geven, dus als ze zeg maar twee of drie collega's een duimpje naar beneden geven, ja, dan gaan we de antwoord optie meteen negeren.”.
In this Dutch quote, [Manager 1] tells about the development of an upvote system, enabling the agents to upvote and downvote the suggested answers. High-rated answers are displayed as the best answer option. Low-rated answer options are automatically removed from the suggestions.
As mentioned in the first paragraph of this case study, the willingness to collaborate with AI tools is present. There are no signs of the inhibitory factor resistance as told by [Manager 1]:
“Niet perse weerstand, wel gewenning. Het is een andere manier van werken.”.
This quote describes the absence of resistance but the presence of habituation. [Manager 1]
describes in his spoken interview the first weeks of the customer support agents working together with the AI tool. After the employee getting used to the collaboration with the tool, the collaboration is going well. According to more general observations of [Employee 5] on the experience with AI, there is no resistance against the use/collaboration of AI tools if the additional value of the tool is evident. This statement is derived from the following quote of [Employee 5]:
“Ik merk weinig weerstand, zodra een AI model kan bewijzen dat het goed presteert en het ons leven makkelijker kan maken is er vaak enthousiasme.”.
[Employee 5] tells in this quote that she does not experience any resistance. Moreover, as soon as an AI model proves its value and can make the employee's work easier, the employee often shows enthusiasm. The quote of [Employee 5] is again an example of the benefit of AI to the Job experience of employees.[Employee 4] describes the general experience of working with AI tools with the following quote:
“everyone thinks it's cool.”.
[Manager 1] also mentions in his interview the coaching role managers play in preventing resistance and safeguarding the correct way of using new technology.
“Managers zullen meer moeten coachen op het accepteren en goed gebruik maken van nieuwe technologie.”
One of the other challenging factors influencing the collaboration between humans and AI is the level of reliability. The reliability of the AI tool used by the support agents in the customer contact center is not yet at the desired level. [Manager 1] emphasizes the learning process of the AI tool. The tool needs to learn from the feedback given by the agents. Sometimes a completely unrelated answer option appears. In the spoken interview, [Manager 1] shares the numbers related to the reliability:
“En nu zitten we tussen de 20 en 40 en we zien dat het steeds meer richting de 40 procent consistent beweegt”
[Manager 1] shows with this quote a percentage of correct answer options between 20 and 40 percent. In the opinion of [Manager 1], a percentage above 50 is desired. Regarding the reliability of AI tools, [Employee 4] consider the stability of the tools as a concern. In his view, the stability issues are the consequences of ABN AMRO being new in the AI field. This observation is shared by [Manager 1], [Employee 5] stresses the importance of safeguarding the quality and stability of AI in processes heavily depending on it already.
The perceived team productivity of the support agents is expected to increase based on the observations of [Manager 1]. The collaboration with the AI tool enables the agents to respond faster. This observation is illustrated by [Manager 1]. with the following quote:
“De medewerker is minder tijd kwijt aan het bijeenzoeken van informatie om het antwoord te complementeren (tel. nrs, links, mails adressen, processtappen).”
This Dutch quote describes the reduction in time an employee has to spend on collecting the correct information, like phone number, links, e-mail address, and process steps.
Next to a quicker response, the introduction of the tool should also lead to an increase in the quality of the provided answers. In the long run, the AI tool should focus on answering the more straightforward questions, and the agent should focus on the more complex issues as explained by [Manager 1]:
“Wat wij eigenlijk willen is dat de agents minder tijd kwijt zijn aan gewoon hele simpele antwoorden … dan kan het model gewoon prima even antwoorden en dat ze meer tijd vrij krijgen om echt complexe problemen te helpen oplossen.”
In this Dutch quote, [Manager 1] expresses the wish that employees spend more time on complex issues and leave the easier problem for the AI tool.
The division of tasks leads to a higher overall quality, according to [Manager 1]. In a year, [Manager 1] hopes to be able to share the results. For example, a reduced waiting time in the queue. The interview of [Manager 1] concludes with the following outlook on the contribution of AI: Eventually, the collaboration with AI leads to ABN servicing more clients with the same number of support or ABN AMRO servicing the same number of clients with fewer support agents. Regarding perceived team productivity, [Employee 4] observes a change of focus of employees to more complex processes after the deployment of AI. [Employee 5] has not made an explicit claim about productivity but stresses the benefits of AI. In her opinion, AI reduces the time of the different processes and enables processes to be executed on a large scale.
The future deployment of AI within ABN AMRO is increasing. All the interviewees expect more presence of AI in the coming years. [Employee 4] thinks the development team will develop more AI-tools/ML models, provided that the development environment is improved. He is advising the bank to allow faster deployment and testing of the different models. Regarding the AI tools deployed in customer support, [Manager 1] expects more autonomous working AI tools within five years. Finally, [Employee 5] foresee an increased deployment of AI if the quality of the AI models improves.
The application of different AI tools/ML models within Rabobank is the subject of the third case study of this research project. An example of one of the various applications discussed in this case study is ML models used to assess the risk of mortgage or loan clients. Further, the risks and the benefits of working together with AI are discussed. The information about the different applications of AI and the related experiences is gathered in an interview with the manager of a data science department. This department is both involved in developing models and training end-users. To refer to the statements and experiences of the manager, the reference [Manager 2] is used.
The contribution of working with AI is significant. [Manager 2] thinks the great benefit of AI above humans is the ability to quickly process large amounts of data (personal communication, April 9, 2021). In the spoken interview, [Manager 2] endorses this benefit with the next quote:
“Ja, mensen zijn best wel slim en heel goed, maar niet zo goed in heel veel repetitief werk op volume”.
In this Dutch quote, [Manager 2] stresses the inability of humans to do high-volume work on a repetitive basis. Another benefit is to leverage ML models to create new products or services.
[Manager 2] proudly tells about the new household expenses feature in Rabobank’s app
powered by an in-house built ML model. This new feature provides both financial insights to the client and the internal employees. For instance, these financial insights support the credit risk department in assessing the financial situation of a client.
Before the collaboration between AI and the employee starts, it is necessary to train and guide the Rabobank employee as told by [Manager 2]:
“Nee, het vergt absoluut begeleiding en training. Ik denk over het algemeen in dit gebied waar mensen vrij onbewust, onbekwaam in zijn. Daarnaast wordt de hele samenwerking met AI onderschat”.
[Manager 2] gives this answer to a question about the necessity of training to stress his concern about the unaware incompetence of employees to work with AI and the underestimation of the collaboration. In the view of [Manager 2], training the employees must prevent underestimating the collaboration with AI from happening. Furthermore, it also prevents the employee from being unable to apply the different AI tools. The level of digital skills of employees within Rabobank ranges from digital illiterate to digital experts, as described by [Manager 2]:
“Al zijn er ook mensen die wat natuurlijker met een soort slimmere tools werken … En er is groep die je die je nog moet vertellen wat er gebeurt als je op de rechtermuisknop klikt.”.
In this Dutch quote, [Manager 2] describes the different levels of digital skills of Rabobank employees. The digital skills of the workforce of Rabobank vary from employees who work naturally with intelligent tools to employees who don’t know what happens if they click the right mouse button. The requisite skillsets of the employee differ significantly between the different AI tools and ML models. However, it is always beneficial if an AI tool is designed to have a low usage threshold, according to [Manager 2].
The collaboration between humans and AI is not inhibited by resistance among the employees of Rabobank. Yet, employees experience the feeling of incomprehension towards Artificial Intelligence sometimes. Following the observations of [Manager 2], many employees