• No results found

Technology Assessment in the Context of Responsible Research and Innovation: Criminal Abuse of Blockchain

N/A
N/A
Protected

Academic year: 2021

Share "Technology Assessment in the Context of Responsible Research and Innovation: Criminal Abuse of Blockchain"

Copied!
20
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Technology Assessment in the Context of

Responsible Research and Innovation:

Criminal Abuse of Blockchain

Author: Dennis Ekkelenkamp

Student number: 11888695

University of Amsterdam Faculty of Science

Thesis Master Information Studies: Business Information Systems Final version: 18-07-2018

Supervisor: dr. Vanessa Dirksen Examiner: drs. Toon Abcouwer

Abstract Technology assessment (TA) has emerged in the 1960s as a way of assessing the total potential societal impact due to emerging technologies. TA was used as an instrument by governments to create new policies for emerging technologies. TA has more or less successfully been used to assess societal impact of emerging technologies. However, with the introduction of responsible research and innovation, some of TA’s problems became more pronounced, especially the vagueness with which soft impacts are assessed. Soft impacts are difficult to assess because they don’t have a direct causal relationship between technology and negative impact, resulting in fuzzy assessments. A potential solution to this problem is the concept of affordances, which describes the relationship between technology, actor, and impact. To discover why the lack of a causal relationship results in fuzzy assessments and if affordances could indeed be used as a solution to the soft impact problem in TA, interviews were held with various experts in three different fields and corroborated by a document analysis of contemporary TA reports. I argue that the concept of affordances can potentially be used as a solution to the soft impact problem by giving an empirically grounded basis for scenarios.

Keywords. Technology assessment, responsible research and innovation, soft impact, affordances

(2)

1. Introduction

New opportunities arise for criminals to continue and improve their illicit business due to the increasing popularity of blockchain applications. The global trade of illegal goods and substances may benefit from some of blockchain’s core aspects, most prominent of these is the ability to provide a high degree of anonymity for their users, making them difficult to trace. While the original goal of blockchain was to make fast and secure transactions across a decentralized network, criminals saw an opportunity to exploit blockchain for their illicit businesses.

Blockchain is not the only technological innovation where this has happened, there is a tradition in technological innovations where they often come accompanied with negative interpretations. Groundbreaking new medicine can later become addictive drugs, the yellow dye called TNT later became a main ingredient in TNT explosives, and the discovery of nuclear fission became the base for the development for hydrogen bombs. These discoveries or inventions in itself were harmless, but eventually used for ill purposes. The negative impact on society due to emerging technologies caused technology assessment (TA) to emerge. TA is conducted to review emerging technologies’ total impact on society, ranging from ecological, economical, and ethical perspectives which is then used by governments to create new policies. 1.1. Problem statement

TA has been around since the 1960s, and has more or less successfully been used to assess the impacts of emerging technologies (Swierstra, 2015). An example of the effects that TA can have on society is the abolishment of research on the Boeing 2707 by the United States’ government. The 2707 was supposed to become the competitor of the Concorde, but the research was cancelled after it became clear that the aircraft would have a strong negative environmental impact (National Research Council, 1975). This is one example that shows how ‘successful’ TA can result in a positive impact on society. However, TA is a lot more difficult in the case of soft impacts, such as in assessing potential criminal abuse of emerging technologies. Soft impacts are impacts that do not have a direct causal relationship between technology and negative impact, which makes these soft impacts difficult to assess. (Van Lente, Swierstra & Joly, 2017). The soft impact problem almost always occurs when assessing criminal usage of technology, because criminal usage of technology is not a causal relationship but is co-produced between technology and actor. Blockchain, for instance, has over the past decade been transformed into a criminal’s playground, processing over 70 billion dollars’ worth of criminal transactions, despite not being an inherently criminal technology(Foley, Karlsen & Putniņš, 2018). Blockchain has allowed for new criminal business opportunities like crime-as-a-service models and easy international transactions. With new technological innovations happening constantly, it is important to account for criminal abuse of technology as part of TA.

With the recent introduction of responsible research and innovation (RRI) an attempt is made to rethink how to conduct TA. Whereas originally the result of TA was used to make new policies, with RRI, the results of a TA are used to actively steer technological innovations in desired directions to improve the positive impact on society, and to lessen the negative impact on society. To accurately steer technology

(3)

RRI emphasizes the predictive function of TA and has thereby enlarged the soft impact problem. Since it is difficult to accurately assess risk of potential criminal abuse of technological innovations, it is also difficult to steer technological innovations. It is therefore important to look at ways of solving the soft impact problem.

While TA focusses on societal impact as a whole, looking at both the positive and negative sides of new technologies, this thesis will take a closer look at the negative societal impact, namely, the criminal abuse of new technologies. In particular the potential criminal usage of technologies is chosen as a topic, since the soft impact problem occurs here often. The focus of this thesis lies on blockchain technology because of its popularity among criminals, making it a good subject to assess the soft impact problem.

The related research question posed is:

How should contemporary technology assessment change to better assess the risk of criminal abuse of emerging technologies?

1.2. Objective

By answering this question, this research aims to improve contemporary TA by looking for a potential solution for the soft impact problem. This is achieved by looking into the way the concept of affordances may improve TA on criminal abuse of technology.

The remainder of this thesis is structured as follows: First a literature review is provided on current TA methods, the connection between RRI and TA, and how the concept of affordances might improve contemporary TA. Next, the applied methodology is discussed. The results and discussion sections will subsequently show how the concept of affordances may potentially be used to better conduct TA of criminal use of technologies. Lastly, a conclusion is provided in which the research question is answered.

2. Literature review

The literature review will start with a definition of TA and is followed by the relationship between TA, RRI, and contemporary problems, and finally the concept of affordances and its potential to improve contemporary TA on criminal use of technology.

2.1. Technology assessment

TA emerged in the 1960s, primarily in the United States, as a method to explore societal issues surrounding the application of technologies (Banta, 2009). TA during this time was technologically deterministic, and served as an early warning system for policy makers to anticipate negative consequences of technology(Grunwald, 2014). Later, in the 80s and 90s, TA gained traction throughout Europe which led to the creation of the European Parliamentary Technology Assessment network(EPTA) (Nazarko, 2017; Grunwald, 2015). During this time a paradigm shift occurred and TA went from technologically deterministic to socially constructivistic; TA was now

(4)

practiced in pursuit of shaping technology (Bijker, et al, 1987; Grunwald, 2014). EPTA advises parliaments on societal, economical, and environmental impact of new

Methods (Tran & Daim, 2008) Methods (Decker, 2011)

Cost benefit analysis Decision analysis Decision analysis

Delphi / expert interviews

Environmental concerns and integrated TA Impact analysis

Information monitoring Measures for technology Roadmapping

Scenario analysis

Structural modeling and system dynamics

Cost benefit analysis Delphi

Discourse analysis Ethical analysis Expert interviews Material flow analysis Modelling and simulation Scenario modelling/workshops Systems analysis

Trend extrapolation Value analysis technologies.

In the broadest sense, TA is about how technological advancements change the world we live in. TA is, however, a generic term. Banta (2009), describes TA as a form of research that aims to examine the long- and short-term consequences of technology. Consequences in this case relating to, for example, ethical, societal, and economical spheres. While Decker (2011), defines TA as a “scientific, interactive, and communicative process, which aims to contribute to the formation of public and political opinion on societal aspects of science and technology” (p.14). More of these definitions exist, and are often based on the author’s personal questions regarding (future) technologies (Decker, 2011). In the context of this research the focus lies on the negative impact emerging technologies can have on society, specifically criminal abuse of technological features.

There are several methods to conducting TA and new methods are continuously being tested. A TA method is chosen depending on the desired outcome, which can be products, future scenarios or policy debates for instance (Decker, 2011). Tran & Daim (2008) conducted a taxonomic report of the various methods used in TA in the past decades; there are two distinct domains for technology assessment: the public decision-making domain and the business & non-governmental domain. The former has been around since the 1969 establishment of the Office of Technology Assessment in the United States’ congress, and was the only domain up until 1995 where TA research was conducted. After the abolishment of the Office of Technology Assessment in 1995, TA started getting more popular in Europe and research was conducted by independent research institutions (universities) and businesses, instead of governmental offices(Tran & Daim, 2008). This shift in research institutions caused the definition of TA to

(5)

become more ambiguous, and marks a shift from the original meaning and conventional application of TA.

Table 1 shows the overarching methods used in TA that where found in the taxonomic report by Tran & Daim (2008), and by Decker (2011). To find out which of the above methods are most commonly used in TA and in assessing soft impacts, interviews and a document analysis will be conducted.

There are however certain criticisms to TA. Namely that predicting societal impact by extrapolating data is over-simplistic, that by using experts as a source of data you might get biased information, and potential negative societal impacts are often described too vaguely (Joss & Belluci, 2002; Van Lente, Swierstra & Joly, 2017 ). As a response to these limitations the concept of responsible research and innovation emerged to improve TA.

2.2 Responsible Research and Innovation and contemporary problems

Responsible research and innovation (RRI) is a concept that gained popularity over the last decade and calls for a greater public engagement for science and technology. RRI gained traction within the EU to cope with apparent limitations of policy making for ethically problematic technological advances. With RRI, the traditional view of ‘scientific freedom’ is challenged and replaced with a promise for innovation that results in positive societal impact (Owen, Macnaghten & Stilgoe, 2012).

Von Schomberg (2012) proposed the following definition for Responsible Research and Innovation: “Responsible Research and Innovation is a transparent, interactive process by which societal actors and innovators become mutually responsive to each other with a view to the (ethical) acceptability, sustainability and societal desirability of the innovation process and its marketable products (in order to allow a proper embedding of scientific and technological advances in our society).” (p.9).

With the concept of RRI, an attempt is made to change TA from a passive solution to an active solution by emphasizing TA’s ability to shape technology (Von Schomberg, 2013). This comes with new challenges however. In the context of nanotechnology, where RRI is practiced most until now, responsible development is characterized as the balancing act of maximizing a technology’s positive impact, and minimizing the negative (Fleischer, Decker & Fiedeler, 2005). RRI involves assessing potential societal implications, a commitment to developing the technology in a way so it fits societal needs best, and making reasonable efforts to minimize negative consequences (Grunwald, 2014). In sum, with RRI an attempt is made to make TA better at shaping technology in desired ways, to better fit society’s needs.

There is however one main shortcoming of TA in the context of RRI; the negative impact on society as described in TA reports are often too vague due to soft impacts (Van Lente, Swierstra & Joly, 2017). Hard impacts are described very clearly in TA reports. Hard impacts are impacts that are quantifiable, where there is a direct causal link between the technology and impact, and the negative impact is clear and noncontroversial. The example of the Boeing aircraft in the introduction is an example of a hard impact: quantifiable impact in the form of an amount of pollution, a direct link between the aircraft and the negative impact (pollution), and a clear and noncontroversial impact as pollution is considered ‘bad’ in society. Soft impacts differ from hard impacts in that the impacts are qualitative rather than quantitative, there is no direct causal link between technology and negative impact, and the negative impact is controversial. These soft impacts are often ignored in TA reports as these impacts are

(6)

regarded as too ‘fuzzy’ to take seriously (Swierstra, 2015). Especially the lack of a causal link between technology and impact makes soft impacts difficult to assess. This importance of a causal link stems from the perspective of technology as passive tools that if used as intended should bring about desirable consequences. However, technology is not always used as intended which may bring about negative consequences (Van Lente, Swierstra & Joly, 2017). A potential solution to the problem of not having a direct causal link due to actors misusing technological innovations can be found in the concept of affordances. Before exploring affordances as a potential solution, the soft impact is shown in relation to blockchain technology.

2.3 Blockchain technology and soft impact

Blockchain technology first emerged in 2008 with the anonymous Satoshi Nakamoto’s paper called ‘Bitcoin: A peer-to-peer electronic cash system.’. Blockchain is a decentralized public ledger which contains all events that have been executed among participating parties. Those events, often transactions, are verified by a majority of the participants of the blockchain (Crosby, et al., 2016). What makes blockchain a unique innovation is that the contemporary digital economy for a large part relies on some trusted third party. With blockchain this trusted third party, like a bank, is not necessary anymore. According to Brown (2016), this absence of a third party does not inherently cause criminal activity, but blockchain is extensively used by criminals. The author explains that blockchain applications relates to crime in the same way physical money does; there is nothing inherently criminal to it.

While blockchain technology is perhaps not inherently criminal, one study found that up to half of all bitcoin transactions are conducted for criminal purposes, causing around 72 billion dollars of illegal activity per year (Foley, Karlsen & Putniņš, 2018 ). This discrepancy between the fact that blockchain and its application are not inherently criminal, and the fact that because of blockchain there is a lot of criminal activity, suggests that the criminal impact of blockchain technology is co-produced by its users, meaning the negative societal impact due to crime shows signs of being a soft impact.

In sum, TA in the context of RRI has a problem with too vaguely describing potential negative (soft) impacts on society and blockchain shows signs of having a soft criminal impact on society. In the next part the concept of affordances is explained and how it might contribute in solving the soft impact problem in contemporary TA. 2.4 Affordances

The term “affordance” was originally coined by psychologist Gibson (1979), and may be seen as that which the environment offers an individual. This is a broad definition and has two distinct movements. Gibson (1979) noted that an affordance exists independent from an observer’s ability to perceive it. An affordance in this case is a sort of ‘use case’ that an object offers an individual within the physical limits of the individual. For instance, a chair affords a person to sit, while the same chair affords a rodent shelter. Norman(1990) changed Gibson’s definition to include a difference between perceived and actual affordances of an object. In Norman’s definition the observer defines the affordances of an object (instead of an object having inherent affordances) by using his past experiences. Norman also notes that designers can make affordances that suggest what they are used for: a big button suggests pressing, a door

(7)

handle suggest turning and pulling, etc. Some affordances are more ‘noticeable’ for their specific use than others, and designers can make use of this by making sure affordances suggest the proper use. As shown in table 2, the two different interpretations of affordances both try to explain the relationship between an environment, an actor, and outcome.

Affordances according to Gibson Affordance characteristics

Offerings in the environment in relation to capabilities of an actor Independent of the actor’s ability to perceive

Binary existence; affordance does or does not exist Affordances according to Norman

Affordance characteristics

Perceived affordances don’t necessarily actually exist

Affordances suggest how they should be used; they have ‘clues’ Can depend on actor’s past experiences

Table 2: Affordances according to Gibson & Norman (Adapted from McGrenere & Ho, 2000)

2.5 Relevance of affordances for improving TA

The concept of affordance, as described earlier, may have potential for improving the way in which TA of criminal use of emerging technologies is conducted. As discussed before, the importance of a causal link between technology, actor, and impacts stems from the perspective of technology as passive tools that if used as intended should bring about desirable consequences, and if misused bring about negative consequences (Van Lente, Swierstra & Joly, 2017). This causal link is however not present in the case of soft impact. Affordances can potentially make this causal link become clearer by describing how an environment (=technology) provides actors (=criminals) with the tools to achieve a certain purpose (=negative impact). Affordances in this case could be seen as a lens through which you look at potential negative features of technologies. With affordances a new method may be created solely for assessing soft impacts, potentially leading to clear assessments instead of fuzzy assessments.

In sum, TA in the context of RRI suffers from the problem of vaguely describing negative societal impact due to a difficulty in assessing soft impacts. Criminal usage of technology is almost always a soft impact, because there is no direct causal relationship between technology, actor and negative impact. Blockchain technology shows signs of potentially suffering from this problem too; intrinsically blockchain is not a criminal technology, but it does account for over 70$ billion in criminal transactions (Foley, Karlsen & Putniņš, 2018). A potential solution to the soft impact problem is the concept of affordances. Affordances are a potential solution to this problem because with affordances you can describe a relationship between technology, actor, and negative impact, which is what is missing in soft impacts.

To see how the soft impact problem occurs in practice, why the fact that there is no causal relationship in soft impacts lead to fuzzy assessments, how TA expert assess criminal usage of blockchain, and to find out if affordances could indeed be used as

(8)

solution to the soft impact problem, interviews and a document analysis are conducted which will be covered in the next chapter.

(9)

3 Methodology

To answer the research question: “How should contemporary technology assessment change to better assess the risk of criminal abuse of emerging technologies?”, interviews and a document analysis are conducted.

3.1 Research approach

To see why the fact that there is no causal relationship in the soft impact problem leads to fuzzy assessments, and to see how TA experts are assessing blockchain technology with regards to criminal usage of the technology, interviews are held. The reason for using qualitative research methods, and in particular interviews, is to get a qualitative understanding of why TA reports are made the way they are and if they indeed have a problem with assessing soft impact. To corroborate these interviews a document analysis of TA reports on criminal use of technologies is held to uncover if they support or disprove the data from the interviews and also feature problems related to soft impacts. This way of triangulating data gives a greater confidence and credibility of the findings (Bryman, 2016; Bowen, 2009).

3.2 Data collection

Semi-structured interviews were held to gather the necessary data. Semi-structured interviewing is the chosen method since this method of interviewing allows for predetermined questions that will extract important data, while interviewees are at liberty to add information that they might deem important. This way important information can still be gathered even if it is not specifically asked for by the interviewer. The aim of the interviews is to uncover how negative impact of technologies is assessed, and if there indeed is a problem of assessing soft impacts. Below in table 3 is an overview of the various experts that are interviewed. These experts where chosen because of their relevance to this research. TA and risk assessment experts were chosen because of their expertise with assessing potential negative impact relating to technology, blockchain experts were contacted because of their experience with blockchain and its potential for good and ill, and police enforcement experts where contacted for their knowledge about criminals and how they use technology for their illicit business.

Field Function

Technology assessment Researcher

Technology assessment Researcher

Blockchain Researcher

Blockchain Researcher

Risk management Manager

Risk management Manager

Police enforcement Detective

(10)

To corroborate the interviews a document analysis was conducted. A document analysis is commonly used to add context to the research, provide additional questions to be asked, supplementary data, and verification of findings (Bowen, 2009). In the case of this research the document analysis mostly served as a way to provide additional data and verify the findings from the interviews.

Two kinds of sampling can be used for selecting documents; random and purposive (Flick, 2014). Purposive sampling is used in this thesis to make sure the documents can provide the data needed to either corroborate or disprove the data provided through interviews. Specifically homogeneous sampling is used, which means samples are chosen based on a particular ‘feature’ that all samples have in common (Flick, 2014). In the case of this thesis that means only selecting documents relating to the assessment of technology and (future) crime. The selected documents are displayed in the table below and are freely available online.

Title Source

Internet Organised Crime Threat

Assessment 2011 Europol

Internet Organised Crime Threat

Assessment 2014 Europol

Internet Organised Crime Threat

Assessment 2015 Europol

Internet Organised Crime Threat

Assessment 2016 Europol

Internet Organised Crime Threat

Assessment 2017 Europol

Een nooit gelopen race Rathenau Institute Technology Assessment Practices in

Europe

Rathenau Institute Waardevol digitaliseren Rathenau Institute Strengthening Technology Assessment

for Policy-Making PACITA

TA Practices in Europe PACITA

Crime Risk Assessment International Compliance Association Critical Infrastructure Protection United States Government

Accountability Office Cybersecurity Actions Needed To

Strengthen U.S. Capabilities UnitedAccountability OfficeStates Government Cyber Crime Assessment 2016 National Crime Agency

National Strategic Assessment of Serious and Organised Crime 2018

National Crime Agency

Table 4: Documents

The above documents were chosen based on their relevance to the research question. All documents look solely or partially into criminal use of technologies, most of which also pay specific attention to blockchain technology. Documents were chosen since the emergence of RRI in 2011, to make sure the documents are still relevant to contemporary TA and this research.

(11)

3.3 Analysis

The interviews were recorded and subsequently transcribed in preparation for analysis. According to Life (1994), qualitative data analysis begins with a complete set of data in the form of text, like transcripts. The text is then completely read after which you read it again and select text for special attention. These fragments of text are then coded and grouped into different categories. These codes allow a researcher to look for important themes. This is usually done with the aid of software, in the case of this research Nvivo was used. The coding scheme created in this research is included in appendix 1.

To analyze the documents, the same codes that were created during analyzing the interviews were used. This was done to corroborate or disprove the findings from the interview with the data from the documents. This is a common practice; document analysis is often conducted with predefined codes from interviews for the purpose of verification (Bowen, 2009).

4 Results

The results relevant to the research question are displayed here. The main method for data gathering, interviews, will be covered first, followed by the results from the document analysis. Both sections are structured according to the categories found in the analysis of the data, starting with the most significant findings. In the discussion following the results, an interpretation of the data is given and how it may answer the research question.

4.1 Interviews

4.1.1 Quantifiable impact

The TA experts mentioned that potential negative societal impact of emerging technologies is mostly measured through numbers. For instance, how much economical damage will a technology potentially produce, or an amount of pollution due to new technological innovations. In determining which risk are most important to manage, the risk managers look at chance of occurrence, time it would take to manage negative outcome, and potential loss of money if the risk would occur. Based on the numbers the managers judge if the proposed risk are actually risks, and then decide to mitigate, avoid, or accept these risks. When asked how risks are assessed that are difficult to quantify, one interviewee said that “(…)that’s the risk of risk-assessment, our methods are not perfect but usually get the job done. Numbers mean more to people, qualitative risk descriptions are often seen as less important. “How much does it cost? What is the chance?” Are questions we often get.” The TA experts noted that often quantifiable risks are all that is needed for a proper TA report. There is not always a need to go beyond the numbers. In ethically difficult technological innovations, like GMOs, TA reports are much more difficult to produce, since numbers mean less here. The same goes for criminal assessments: “(…)there are parts that can be quantified, such as the current amount of money laundered through blockchain, but assessing possible future criminal use of technology is very difficult. We don’t want to invent arbitrary scenarios in the hope of uncovering potential negative societal impact due to technologies, on what basis do we make these scenarios? ”.

(12)

One TA expert, who is in the midst of conducting a TA on blockchain technology, had never heard of the concept of affordances. When asked how the possible negative societal impact relating to crime was assessed as part of his TA report, he said: “(…) we try to uncover how a technology can be used for ill by looking at the technological features. In the case of blockchain the absence of a third party overlooking transactions is a very enticing feature for criminals.”

A blockchain expert said that risk in the case of blockchain, whether that’s economical risk, criminal risk, or ethical risk, is assessed in the ‘traditional’ sense with chance of occurrence and impact. “We look at the capabilities of different blockchain applications and how they may impact certain businesses. Is it likely that blockchain will be used for voting in elections for example? And what would that mean for society?” The law enforcement expert said that usually an investigation into technology only starts when there is ‘blood on the street’. “Law enforcement is expected to keep society safe, very visible crimes will get priority over invisible ones. Drugs dealers operating on the street are more important to deal with in the public eye than drug dealers dealing via bitcoin and dark webs.” Only when it becomes a truly big issue, and society deems it important too, will police enforcement look more seriously into the issue of blockchain related crime. Assessments by law enforcement are mostly to analyse the status quo.

4.1.2 TA methods & fuzzy assessments

When asked which methods are commonly used in making TA reports, the TA experts noted that: “we always use literature reviews and expert interviews as methods to get a basic understanding and for context. After which, depending on what kind of assessment, we employ methods ranging from cost-benefit analysis, to scenario modelling and impact analysis ” The risk managers used solely methods relating to chance and impact: “(…) I assess risk in a business setting, which means that I really only look at a chance of something happening, and how much of an impact that will bring to business processes.” When asked if different methods are used for assessing different kinds of impacts, like the soft and hard impacts, experts said “we try to steer away from addressing very qualitative impacts, as they are very difficult to assess ”.

4.1.3 Innovative criminals

When asked how risk assessment relating to criminal usage of emerging technologies could be improved, one blockchain expert noted that ‘Innovation is often thought of as something that happens solely in academia or business, but innovation also happens in the criminal world. Bitcoin in itself is not too interesting for criminals, but coupled with the TOR network, and criminally developed ‘bitcoin tumblers’ the cryptocurrency can be exploited for illicit business. Criminals saw an opportunity and figured out how to exploit the technology to meet their interests. Law enforcement agencies only started investigating exactly how cryptocurrencies are misused when it was already a big problem. If they were more proactive in predicting how blockchain’s features might be coupled with other technologies, like TOR, then law enforcement would perhaps not be playing catch-up like it does now.’. The expert works closely with law enforcement and conducts research into criminal usage of blockchain. When asked how the assessment should have taken place back when blockchain technology was first introduced, he said “(…) it is very difficult however to predict exactly how a technology might be abused

(13)

by a criminal, since there are endless ways in which you could combine technologies and practises”.

4.1.4 Blockchain impact

When asked why blockchain technology is used for criminal purposes so much, most interviewees agreed that the anonymity that some blockchain applications, like Bitcoin or Monero offer, is the main attraction for criminals. One blockchain expert noted that ‘(…) the perceived anonymity is the main attraction for criminals to use Bitcoin. You could say however, that Bitcoin is pseudonymous instead of anonymous. For true anonymity, blockchain applications like Monero or Zcash are used. Criminals continuously look for ways to improve the technology’s ability to hide them from law enforcement. Another reason for the popularity of blockchain applications according to blockchain and law enforcement experts are the relatively easy international transactions: ‘(…) not having to worry about international transactions is a big deal for criminals. this is perhaps an effect of the anonymous properties of Bitcoin or Monero. These currencies are not merely used for anonymous payments, but also the transfer of value across border without supervision of authorities’.

And what is perhaps the most important factor in the popularity of blockchain applications like bitcoin or Monero is that they are very easily used: (…) the entrance barrier to cybercrime has been lowered substantially by cryptocurrencies. People can now simply buy ransomware instead of having to produce it themselves”.

4.2 Documents

The results from the document analysis are displayed here, structured according to the categories found in analysing the interviews.

4.2.1 Quantifiable impact

The experts in the interviews noted that impact is often expressed though numbers. Looking at various TA reports, the same phenomenon can be seen. TA reports generally give a very thorough analysis of how a technology is currently being abused by criminals, and often come to the same conclusions as other TA reports regarding criminal usage; in all analysed documents containing blockchain technology, bitcoin was seen as a key enabler of dark web trade. The assessments also succeeded in explaining how exactly criminals use different combinations of technologies like the TOR network, VPN services, and bitcoin tumblers to conduct their illicit businesses. However, most reports lack a distinct predictive function, which is an important part of TA in the context of RRI.

Reports by Europol are generally more in-depth relating to criminal usage of technology then their other TA counterparts. Europol’s reports are extensive and paint a broad picture of the status quo. The predictive function is however limited. This becomes clear when looking at reports by Europol from 2011 to 2017; the earlier reports pay no attention whatsoever to blockchain, while two years later a sizable portion of the report is dedicated to the technology. This means that Europol excels at thorough analysis of the status quo of criminal usage of technology, but perhaps lacks a thorough predictive function, at least in the case of blockchain technology.

(14)

4.2.2 TA methods & fuzzy assessments

Depending on the source of the reports, not all covered their methods. Documents from the Rathenau institute did mention their methodology, in “een nooit gelopen race” (Rathenau, 2017), expert interviews and a literature study were conducted first, after which a trend analysis was conducted based on existing data to try to predict future impact. In both “Critical infrastructure Protection” (United States Government Accountability Office, 2010), and “Internet Organised Crime Threat Assessment” (Europol, 2017) the applied methods are a document analysis, interviews, and a survey. Other reports show a preference for quantitative methods, saying little about possible future impact. TA reports also feature fuzzy assessments, staying in generalities. An example of this: “It is likely that reporting of fraud will continue to show a steady rise over the next 12 months. The continued push for easily accessible online services will continue to provide opportunities for fraudsters” (National crime agency, 2018).

4.2.3 Innovative criminals

The blockchain experts said during the interviews: “(…)but coupled with the TOR network, and criminally developed ‘bitcoin tumblers’ the cryptocurrency can be exploited for illicit business. Criminals saw an opportunity and figured out how to exploit the technology to meet their interests.”, documents from Europol, the national crime agency, and the international compliance association confirm this notion. So called cross-cutting threats are assessed in the reports by investigating how criminals combine different kinds of technologies to meet their interests. The reports specifically mention the TOR network and VPN services as prominent technologies that criminals employ in combination with blockchain.

4.2.4 Blockchain impact

The reports, like the experts, also show that the entrance barrier to cybercrime has been lowered. To remain completely anonymous however still takes some know-how (National crime agency, 2018). Regarding potential future criminal usage of blockchain, TA reports tend to remain in generalities. For instance, the anonymous properties of blockchain are considered the main attraction to criminals, on the basis of which reports suggest that in the future cryptocurrencies like Monero or Zcash will become even more popular. But more ‘outside-of-the-box’ predictions are not made, which is perhaps necessary in the case of soft impacts.

(15)

5 Discussion

In this section the results are discussed to see how they may answer the research question.

5.1 Affordances in contemporary TA

The results show that there is indeed a problem related to soft impacts in contemporary TA. Perhaps most importantly, in contemporary TA there is made little distinction between hard and soft impacts, results in using the same methods for both, or soft impacts being avoided if possible. TA is precise in assessing negative hard impacts. When assessing soft impacts, like the co-production of crime by technology and criminal, TA is much more vague and imprecise, staying in generalities. As discussed in the literature review, soft impacts are not taken seriously in contemporary TA (Swierstra, 2015). The results however show that the soft impacts in criminal TA are taken seriously, but they are indeed fuzzy. One TA expert noted that he did not want to create arbitrary future scenarios to possibly reveal negative societal impact due to technology, and would rather stay with his current methods. There is a certain willingness to assess these soft impacts but not a good way of doing so with current methods.

The document analysis revealed that in assessing potential criminal use of technologies as part of TA, there is a distinct lack of coverage on soft impacts. In the TA reports covered in the document analysis, researcher mainly try to quantify the negative societal impact due to criminal usage. This may not be sufficient to cover the load, since there is not always a direct causal link between the technology and the negative impact due to criminals. Looking at blockchain, this phenomenon becomes clearer: by itself blockchain technology does not directly impact society in a negative way, it merely allows for decentralized transactions (Brown, 2016). However, when coupled with other technologies like the TOR network, blockchain technology can be abused for criminal purposes. This co-creation of negative impact between technology and criminal is hard to account for in traditional assessment methods, like the methods used in assessing hard impacts (Swierstra, 2015).

The methods used in assessing hard impacts are also often used in assessing potential soft impact, which can cause the vagueness with which these potential negative impacts are assessed. Instead, in TA, researcher should perhaps aim to use different methods when assessing soft impacts. One proposed method of assessing soft impact is the usage of ‘stories’ or scenarios as Swierstra (2015) puts it. These stories however should be grounded empirically; describing what is technologically possible and how technologies and practices interact (Lucivero, Swierstra & Boenink, 2011). Affordances can enrich these stories by describing the capabilities of technology with regards to certain actors; the relationship between technology and criminal, thereby grounding the stories empirically.

In sum, affordances could potentially be used to get closer to a solution to the soft impact problem in contemporary TA. Through interviews it is discovered that the problems relating to soft impacts are the result of using the same methods for both hard and soft impacts. The fear of making arbitrary stories can be lessened by using affordances to ground the stories empirically by describing the relationship between

(16)

technological features and criminal capabilities. Affordances can also describe the relationship between technology, criminal and impact, which is missing in contemporary assessment of soft impacts. As discussed in the literature review, there are two main interpretation of affordances however. To see how the two definitions of affordances differ in practice when used in assessment for criminal use of technology, and which definitions lends itself the most for a solution to the soft impact problem, the following scenario is applied to the two definitions of affordances based on expert input.

‘A (cyber)criminal wants to sell illegal goods and or substances online. He intends to use an online dark web market to conduct his illicit business. However, he noticed a recent trend in technology; bitcoin. He knows that by combining the anonymous properties of the TOR network, and the decentralized properties of blockchain technology, he can conduct his illicit business without much to fear in terms of repercussions.’

In this example, the combination of TOR and blockchain technology afforded the criminal to conduct his illicit business. In table 5 the differences between the two definitions of affordances are revealed through applying the above example.

Affordances according to Gibson

Affordance characteristics Example Offerings in the environment in relation to

physical capabilities of an actor Blockchain can only be used if the criminalcan physically operate a computer or phone Independent of the actor’s ability to

perceive Blockchain technology can be used for transactions without the criminal’s knowledge

Binary existence; it does or it does not exist The affordances blockchain offers only presents itself in certain cases; when the criminal has no internet connection, blockchain affordances don’t exist Affordances according to Norman

Affordance characteristics Example Perceived affordances don’t necessarily

actually exist Blockchain applications may prove to be less anonymous then it would initially suggest; bitcoin’s ‘anonymity’ is merely pseudonymity

Affordances suggest how they should be

used; they have ‘clues’ Decentralization dictates that there is no central party that observes transactions, which suggest blockchain lends itself for criminal usage.

Depend on actor’s past experiences An experienced cybercriminal can recognize the potential of blockchain technology that an inexperienced (cyber) criminal would have missed

Table 5: Affordances (Adapted from McGrenere & Ho, 2000 )

In the above scenario, Norman’s definition of affordance is more fitting when assessing potential negative impact of technology, since Norman’s definition accounts more for

(17)

individual differences. For instance, a run-of-the-mill (street) criminal probably won’t have the necessary experience to recognize blockchain’s potential for criminal business. When assessing risk, it is important to understand these individual differences. Using Gibson’s definition, blockchain(the environment) affords all criminals (individuals) new ways to conduct illicit business. Using Norman’s definition, only those criminals who are tech-savvy will ‘receive’ the affordances that blockchain technology offers. In assessing potential negative impact of criminal use of emerging technologies, using one definition over the other might result in much different pictures. Gibson’s will produce a higher-risk scenario, since all criminals will have access to blockchain affordances, while Norman’s definition produces a lower risk scenario, since only a select group of criminals will have access to blockchain’s affordances. Accounting for these individual differences in criminals is therefore important when assessing potential negative impact. Another reason to use Norman’s affordances over Gibson’s is that criminals, according to interviewed experts, tend to combine technologies, TOR and blockchain for instance. The clues that affordances give are more in line with this way of thinking, as they suggest how they might be used coupled with other technologies.

6 Conclusion

The aim of this research was to improve contemporary TA relating to criminal abuse of technology with a research question: How should contemporary technology assessment change to better assess the risk of criminal abuse of emerging technologies? An improvement of current TA is necessary since the introduction of RRI has emphasized the soft impact problem that plagues TA. This problem is especially prevalent in criminal usage of technologies, since criminals use technologies for purposes not intended, which means that there is no direct causal relationship between technology and negative impact. Contemporary TA experts mostly use the same methods for assessing hard and soft impacts, resulting in fuzzy assessments of soft impacts. A possible solution to this problem is the concept of affordances which can describe the relationship between technology, actor, and impact, which is what is missing in assessing soft impacts. Through interviews corroborated by a document analysis, it is revealed that experts in relevant fields indeed struggle with clearly describing soft impact with current assessment methods. Using affordances to create empirically grounded stories, or scenarios, can the soft impact problem potentially be solved.

(18)

References

Banta, D. (2009). What is technology assessment?. International journal of technology assessment in health care, 25(S1), 7-9.

Bijker, W. E., Hughes, T. P., Pinch, T., & Douglas, D. G. (2012). The social construction of technological systems: New directions in the sociology and history of technology. MIT press.

Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative research journal, 9(2), 27-40.

Brown, S. D. (2016). Cryptocurrency and criminality: The Bitcoin opportunity. The Police Journal, 89(4), 327-339.

Bryman, A. (2016). Social research methods. Oxford university press.

Crosby, M., Pattanayak, P., Verma, S., & Kalyanaraman, V. (2016). Blockchain technology: Beyond bitcoin. Applied Innovation, 2, 6-10.

Decker, M. (2011). Bridges between science, society and policy: Technology assessment - methods and impacts. Berlin: Springer.

Europol. (2017). Internet Organised Crime Threat Assessment. Retrieved from: https://www.europol.europa.eu/activities-services/main-reports/internet-organised-crime-threat-assessment-iocta-2017

Fleischer, T., Decker, M., & Fiedeler, U. (2005). Assessing emerging technologies —Methodological challenges and the case of nanotechnologies. Technological Forecasting and Social Change, 72(9), 1112-1121.

Flick, U. (2014). An introduction to qualitative research. Sage.

Foley, S., Karlsen, J., & Putniņš, T. J. (2018). Sex, drugs, and bitcoin: How much illegal activity is financed through cryptocurrencies?.

Geert Munnichs, Matthijs Kouw & Linda Kool, Een nooit gelopen race - Over cyberdreigingen en versterking van weerbaarheid. Den Haag, Rathenau Instituut 2017

Gibson, J. J. (1979). The theory of affordances. The people, place, and space reader, 56-60.

Grunwald, A. (2011). Responsible innovation: bringing together technology assessment, applied ethics, and STS research. Enterprise and Work Innovation Studies, 31, 10-9.

Grunwald, A. (2014). Technology assessment for responsible innovation. In Responsible Innovation 1 (pp. 15-31). Springer, Dordrecht.

Grunwald, A. (2015). Technology assessment. In Encyclopedia of Information Science and Technology, Third Edition (pp. 3998-4006). IGI Global.

Harro van Lente, Tsjalling Swierstra & Pierre-Benoît Joly (2017): Responsible innovation as a critique of technology assessment, Journal of Responsible Innovation (LOOK THIS SOURCE UP FOR BETTER APA^^)

Joss, S., and S. Bellucci, eds. 2002. Participatory Technology Assessment: European Perspectives. London: Centre for the Study of Democracy

Life, R. S. (1994). Qualitative data analysis.

Lucivero, F., Swierstra, T., & Boenink, M. (2011). Assessing expectations: Towards a toolbox for an ethics of emerging technologies. NanoEthics, 5(2), 129.

McGrenere, J., & Ho, W. (2000, May). Affordances: Clarifying and evolving a concept. In Graphics interface (Vol. 2000, pp. 179-186).

(19)

National Research Council. 1975. Environmental Impact of Stratospheric Flight: Biological and Climatic Effects of Aircraft Emissions in the Stratosphere. Washington, DC: The National Academies Press. https://doi.org/10.17226/20101.

Nazarko, Ł. (2017). Future-Oriented Technology Assessment. Procedia Engineering, 182, 504-509.

Norman, D. A. (1990). The design of everyday things. New York: Doubleday. Owen, R., Macnaghten, P., & Stilgoe, J. (2012). Responsible research and innovation: From science in society to science for society, with society. Science and public policy, 39(6), 751-760.

Swierstra, T. (2015). Identifying the normative challenges posed by technology’s ‘soft’ impacts. Etikk i praksis-Nordic Journal of Applied Ethics, (1), 5-20.

Tran, T. A., & Daim, T. (2008). A taxonomic review of methods and tools applied in technology assessment. Technological Forecasting and Social Change, 75(9), 1396-1405.

U.S. Government Accountability Office. (2004). Cybersecurity for Critical Infrastructure Protection. Retrieved from https://www.gao.gov/products/GAO-04-321

Von Schomberg, R. (2011). Towards responsible research and innovation in the information and communication technologies and security technologies fields.

Von Schomberg, R. (2012). Prospects for technology assessment in a framework of responsible research and innovation. In Technikfolgen abschätzen lehren (pp. 39-61). VS Verlag für Sozialwissenschaften

(20)

Appendix: Coding schedule

Categories Codes Examples

Difficulties in TA Soft impact

Arbitrary stories

Predictive assessment

Same methods for soft and hard impacts

Qualitative risk assessments are often seen as less important.

We don’t want to invent arbitrary scenarios in the hope of uncovering potential negative social impact

Predicting how criminals might abuse technologies is exceptionally difficult. There are countless of ways of misusing a technology.

Ethically problematic areas are difficult to assess. Quantifiable impact Quantitative preference

Chance & impact

People are interested in the numbers, what will this risk cost us?

We express risk in our business by looking at chance of occurrence and potential negative impact

Innovative criminals Combining technologies

Criminal innovation

Blockchain is not very interesting to criminals in itself. By combining blockchain with tumblers, VPNs, and the TOR networks does it become criminally viable.

Innovation is seen as something that happens solely in academia and business, but innovations happens for and by criminals too

Blockchain impact Anonymity / pseudonymity

International transactions

The main attraction that blockchain offers criminals is the perceived anonymity, or pseudonymity

The ease with which coins can be transacted across borders is also very enticing to criminals.

Affordances No knowledge of the concept

Describing technological features

No, I have never heard of that concept. We do look at what a technology offers a criminal however, we just call it different.

We try to uncover through literature review and expert interviews what a technology offers its users.

TA methods Expert interviews

Cost-benefit analysis

We try to uncover through literature review and expert interviews what a technology offers its users.

(21)

Trend extrapolation

technology both positively and negatively affects society, we for instance say how much economical gain and loss we e

Looking back at previous data we can make predictions about the future

Ad-hoc assessment Blood on the street Police enforcement prioritizes visible crimes Blockchain trends Monero & Zcash

Crime as a service

We see a migration to other currencies that provide an even higher amount of anonymity

Blockchain facilitates criminals to sell services over the dark web

Keeping up Keeping up

supervision

Law enforcement is always a few steps behind, playing catch-up

Who will supervise bitcoin exchange offices? Contemporary TA

practices Discussion

Inclusive research and innovation

Public discussion has becomes more and more important in ethically problematic technological advancements

It is important to get both experts and societal actors together

Referenties

GERELATEERDE DOCUMENTEN

Mogelijk hierdoor zijn de verschillen in de hele najaarsteelt in kg productie, aantal vruchten en vruchtgewicht gering ondanks een verschil van 2800 groeigraaduren tussen de

According to our results, urban sparrows showed higher levels of oxidative damage and higher activity of antioxidant enzymes, but lower antioxidant capacity in comparison with

Platforms and design methods for innovation are sometimes recommended for their potential to create developments that cannot be predicted nor anticipated, which

An empirical study was undertaken, using 71 entrepreneurs in Rustenburg (l\Jorth West Province) and its environs. The conclusions and comparison that are drawn

The argument put forward in this article is that, by following a positive discipline approach, educators can carry out their mandate to promote and observe human rights simply

De in de tabel vermelde aanbevelingen van rassen zijn conform de Aanbevelende Rassenlijst voor Landbouwgewassen 2005; A = Algemeen aanbevolen ras, B= Beperkt aanbevolen ras, N =

Naast de vele lezingen, cursussen, borrels en andere bijeenkomsten, kijk ik met veel plezier terug op de ‘Ameland week’ in 2013, maar ook de publieke optredens die jullie mij

Er werd verwacht dat een hoge score op Motivatie, Attitude tegenover leersituatie en Instrumentele oriëntatie positieve richtingscoëfficiënt zouden hebben en dat prestatie op