Comparing Nudges and Deceptive Patterns at a Technical Level

Abstract

Nowadays, two concepts are widely used to influence users' behavior on websites. These are deceptive patterns and nudges. In the literature, the two concepts are distinguished according to their goals and efects. While deceptive patterns are considered as manipulation of users, nudges have a more positive connotation - they are meant to motivate users to make better decisions. However, looking at these concepts from a technical perspective, the question arises whether they also difer in their realization. Is it possible to automatically diferentiate between them while crawling a web page for deceptive patterns? To answer this question, we developed a methodology that we present and apply in this paper. Furthermore, we show that there is no need to distinguish between the two concepts, because they are implemented using the same techniques.

Keywords

• eol>dark patterns
• deceptive patterns
• nudging

1. Introduction

Our research focus is to automatically identify if website providers use deceptive patterns on their websites. Thereby, one of our goals is to observe and analyze changes of cookie banners in terms of deceptive patterns over time. Consequently, we could use the analysis results, e.g., to determine whether legal regulations have an impact on cookie banners in the EU or worldwide and for other topics relevant with regard to privacy as deceptive patterns are more and more used to undermine the users defense. To achieve our goal, the logical start would be to capture cookie banners and, then, identify if they use deceptive patterns. However, to identify deceptive patterns, we must first understand what a deceptive pattern is. Only then we can choose appropriate technical tools to detect them. Hence, our first research question is not only to understand what a deceptive pattern is, but also to systemize the types of deceptive patterns and to determine whether each type is technically detectable. Furthermore, website providers could also use nudges. At first impression, nudges are a diferent concept than the deceptive patterns. However, if you try to distinguish between the two concepts from a crawler’s point of view, uncertainty arises as to how exactly you should diferentiate between them technically. Consequently, our second research question is about the implementation of deceptive patterns and nudges in websites. Do we have to diferentiate between deceptive patterns and nudges if crawling for them?

To find answers, we developed and applied a methodology for comparing deceptive patterns and nudges. Our results show that there is no diference between deceptive patterns and nudges from the technical point of view. This methodology also applies to upcoming types of deceptive patterns or nudges. Moreover, we provide an overview of currently detectable deceptive patterns and nudges from the perspective of a crawler.

The rest of the paper is organized as follows: In Section 2, we present our system model. Then, in Section 3 and 4, we define the terms deceptive pattern and nudging to understand the concepts behind them. We describe our methodology for comparing those concepts in Section 5. To decide whether both concepts are the same from a technical point of view, we apply our methodology and analyze the results in Section 6. In Section 7, we discuss related work. Finally, we conclude our paper with a brief summary and provide an outlook in Section 8.

2. System Model

Our system is based on the idea to use a crawler to examine websites in the context of deceptive patterns. Such a crawler could be a framework like OpenWPM (Open Web Privacy Measurements). The purpose of this framework is to assist researches and developers in analyzing privacy-related aspects of web browsing and online tracking [ 1 ].

Our crawler mimics a user visiting a webpage, analyzing its source code to extract visible parts, ensuring the data reflects what a user sees for accurate identification of deceptive patterns. We consider only the types of deceptive patterns whose capturing is technically feasible. Hence, deceptive patterns that aim to trigger certain reaction in users through the content of images or texts are out of our scope.

In general, the providers of websites have specific goals. According to these goals, the web designers create user interface designs to guide people’s behavior in digital choice environments [ 2 ]. We refer to such user interface designs from now on just as design. Furthermore, we use the term technique to refer to technical implementations of these designs. For instance, a popup might utilize HTML for structure, Javascript for loading, and CSS for styling.

In summary, the main components of our system are web pages, the techniques used on the web pages, and a crawler that analyses the designs of the web pages. We focus on the designs that are shown to users and are technically identifiable. Thereby, legal and ethical aspects are not part of this work, as they are not relevant to decide if it is technical feasible to identify such designs or not.

3. Deceptive Patterns

As Harry Brignull coined the term dark pattern and later replaced it with the term deceptive patterns, we use his definition. In [ 3 ], he identifies deceptive patterns as manipulations on websites and apps causing unintended actions, like unintended purchases or sign-ups. These patterns are unique from regular design patterns as they manipulate human decision-making weaknesses to prompt user action [ 4 ].

In [ 5 ], Luguri and Strahilevitz show the efectiveness of deceptive patterns. They examined how variations in user interface design, including mild dark patterns and additional click-through screens, significantly increased acceptance of a questionable data protection plan among diferent groups.

Besides the term deceptive pattern, there are other terms with the same or very similar meaning. These terms are: dark patterns, aggressive dark patterns [ 5 ], dark nudging [ 6 ], deceptive content [ 7 ], deceptive dark patterns [ 8 ], deceptive pattern [ 3 ], digital dark nudge [ 9 ], mild dark patterns [ 5 ], sludge [ 10 ]. We consider these terms as synonyms and use only the term deceptive patterns in the rest of the paper.

4. Nudging

Unlike deceptive patterns, there is no standard and precise definition of the term nudge. We adopt Richard Thaler and Cass Sunstein’s interpretation from the updated version of their book, aligning with Meske et al. [ 11 ]. However, we reference the updated edition [ 10 ]. Accordingly, Thaler and Sunstein [10, p. 6-8] describe nudges as choices left open, unobstructed, or unburdened, emphasizing freedom of choice. They further detail that nudges should guide people towards improving their lives, alter behavior predictably without limiting options or altering economic incentives significantly, and be easy and inexpensive to avoid.

Similar to deceptive patterns, there are various terms for nudges that refer to more or less the same meaning. In the following, we list some examples: bright pattern (“privacy-friendly design nudges”) [ 12 ], default nudges [ 13 ], digital nudging [ 2 ], green nudging [ 14 ], nudging elements [ 11 ], nudging mechanisms [ 15 ], nudging techniques [ 16 ], privacy nudges[ 13 ], self-nudging[ 17 ], soft paternalistic interventions[ 18 ]. We treat these terms as synonyms and use the term nudges in the rest of the paper.

5. Methodology

To address our research questions outlined in Section 1, we systematically explore the similarities and diferences between deceptive patterns and nudges using a methodology we developed. This section details each step of our approach.

Step 1. Literature Review We start with the literature review. When searching for relevant papers, we use the following keywords or their combination: dark patterns, deceptive patterns, comparing, comparison, vs, nudges, and nudging. To ensure the relevance of the papers, we set specific criteria: 1. It was published after 2008. The reason for this is that the term nudge was only coined in 2008. In turn, the term dark pattern was introduced in 2010. 2. It contains an explanation of specific designs. 3. The authors consider either deceptive patterns or nudges. 4. The authors compare deceptive patterns and nudges.

We divide the literature review into two parts: the search for papers on deceptive patterns, and the search for papers on nudges. To organize the review results, we use a separate table for each part. These tables consist of two columns: In the first column, we list the name of a design. In the second column, we place a short description of the corresponding design. With a design, we mean either a deceptive pattern or a nudge.

This step serves to get an overview of existing types of deceptive patterns and nudges exist. It results in two tables: the one contains the identified deceptive patterns, the other the nudges. Step 2. Implementation Review Based on the results from the previous step, we examine which technique exist to implement deceptive patterns or nudges in web applications. For this, we analyze the current literature and assess the identified methods for relevance to our project. We consider a technique as relevant if it fulfills the following two requirements: 1. It is applicable to implement a deceptive pattern or a nudge. 2. It can be detected automatically, e.g., with the help of a crawler.

This step is split into two parts: one focusing on deceptive patterns and the other on nudges. Using the tables from Step 1, we identify applicable techniques for each design. To organize the results, we add another column. Thus, we get a table with three columns: the first for the designs, the second for the description, and the third for the technique. Per row, we get for each design an information with which technique it can be implemented.

This step serves to gather techniques that are available to implement the designs collected in Step 1. It results in two extended tables: the one gives an overview of relevant techniques available for deceptive patterns, the other for the nudges.

Step 3. Filtering Due to our requirements in Step 2, it is possible that some designs were not extended with a technique. Therefore, in this step, we analyze the two resulting tables from Step 2. If there is an item without a technique in the corresponding column, we delete this row.

The aim of this step is to get an overview of automatically detectable designs. This step results in the two cleaned tables: the one gives an overview of relevant deceptive patterns (Table 1), the other for the nudges (Table 2). In the following, we consider only the designs that remain after this step.

Step 4. Comparison In this step, we com

pare the deceptive patterns and nudges to each other. For that, we analyze the two tables from Step 3. To store the results of our comparison, we prepare another table (Table 3) with the following columns: Deceptive Pattern, Nudge and Technique. Next, we transfer the columns Deceptive Pattern and Technique from Table 1 to Table 3. For the following analysis, we can ignore Table 1, as the relevant information is now available in Table 3.

Next, for each row in Table 3, we examine the column Technique in the table with nudges (Table 2) to check if there is an item that uses the same technique. Whenever we find a match in Table 2, we mark the corresponding nudge name in bold indicating that we have already considered it. Additionally, we put the name of this nudge in column Nudge in the current row in Table 3. This step is completed after we have considered all rows in Table 3.

With this step, we aim to identify if there are deceptive patterns and nudges that use the same techniques. The result is an overview of designs that might have diferent names, but are implemented with the same techniques.

Step 5. Identifying Diferences In this step, we examine Table 2 and focus on the entries that are not marked in bold. They were not marked in the previous step, because none of the examined deceptive patterns uses the same technique. Therefore, we consider the corresponding nudges as distinguishable by the technique used. For a better overview, we copy the not marked rows into a new table (Table 4).

We complete our comparison of deceptive patterns and nudges when we make a decision about each design we collected in Step 1. In the end, we have two tables: Table 3 with designs that we cannot automatically distinguish, and Table 4 with nudges that we can distinguish Figure 1: Activity Diagram representing our from deceptive patterns by the technique. Based Methodology on the results after Step 5, we should be able to answer our research questions. Our methodology is represented as a diagram in Figure 1.

It must be kept in mind that the results are only a snapshot. Technology is continuously evolving, and new insights into behavioral psychology are possible. Thus, innovative techniques may soon be available to implement old or novel nudges or deceptive patterns. However, with our method, the results from previous iterations can be extended by new elements to analyze the current situation.

6. Analysis of Diferences

In this section, we apply our methodology (see Section 5) to determine if, and if so, what techniques we can use in order to distinguish deceptive patterns from nudges.

Step 1. Literature Review In this step, we identified that Mathur et al. [ 19 ] published a literature review on deceptive patterns. In this paper, the authors organize the deceptive patterns by paper. As a result, they provide a list of deceptive patterns for each paper. Based on our previous research, we can confirm that they list all deceptive patterns that we know. Additionally, they provide a short description for each listed deceptive pattern. Hence, we can use this list for our analysis. We transfer the listed deceptive patterns with their descriptions in a table (Table 1). However, some deceptive patterns are listed in several groups. Thereby, they have diferent names, but a very close description. For example, there is a deceptive pattern ‘Bait & Switch’ with the description “You set out to do one thing, but a diferent, undesirable thing happens instead.” from [ 3 ]. A very similar deceptive pattern with the name ‘Bait & Change’. In [ 20 ], the authors describe it as “A setting or a choice made by the individual produces a diferent result than desired.”[ 20 ]. In such cases, we select only one version and omit others. Additionally, we group the same deceptive patterns together, for example ‘Address Book Leeching’ and ‘Social Pyramid Schemes’, as they both are based on a social factor to obtain contacts or contact lists.

There is a similar work on nudges. Jesse et al. [ 21 ] conducted a systematic literature review on nudges and provide a well-structured list of existing nudges. We transfer these nudges and their descriptions in the second table (Table 2).

Step 2. Implementation Review No existing work catalogs techniques for implementing deceptive patterns or nudges. Therefore, we analyzed web design techniques meeting our criteria: applicability to deceptive patterns or nudges and automatic detectability.

In the following, we list the techniques that fulfill our requirements, and give an example of what they can be used for: audio (a sound to create attention), color (aggressive colors to increase the salience), fonts (text difers from the rest of the text/website), highlighting (changes the background color, putting a frame around text), images (catching the attention of the user), mandatory field entries (field in a form that need to be filled out to proceed), popup (block the website completely, or partially, to urge the user into interaction with the popup), preselected checkboxes (opt-out, preselected options which the user have to deactivate), progress bar (to show the user how close or far he is to the goal).

Another technique which we consider as detectable, but is not implemented with a single technique is nested path, where the user have to click through, e.g., menus. The last technique that we consider relevant is wording. It is not really a technique in our definition, but it is an important component of many deceptive patterns and nudges, e.g., when communicating the behavior of others.

Our focus on basic techniques excludes methods like blinking or rapid color changes, as these are combinations of color and highlighting—techniques already accounted for in our analysis.

Having the list with techniques, we extend Table 11 with the column Technique. After that, we verify row by row which technique can be used to implement the deceptive pattern in the current line. If we do not find an appropriate technique, this cell remains empty, and we proceed with the next row. We do the same for the table with nudges (Table 22).

Step 3. Filtering In this step, we check Table 1, scrutinizing the column Technique for empty cells. Empty cells indicate that the corresponding deceptive pattern lacks a technically detectable implementation method. Consequently, we delete rows with empty Technique cells. In general, this has afected deceptive patterns that are based on complex ideas and cannot be implemented with basic techniques. For instance, the collective term ‘sneaking’ is too broad in its descriptions as “an attempt to hide, disguise, or delay the divulging of information that has relevance to the user” [25]. We cannot

1Due to the size, we show Table 1 only after Step 3. 2Due to the size, we show Table 2 only after Step 4.

Address Book “A service provider ofers users to upload or import their address books

Leeching, Social to connect with known contacts on that service.” [22]

Pyramid Schemes

Bad Defaults[22], “[. . . ] usually manifests as a default choice [. . . ]; however, this preselected

Default choice is often against the user’s interests or may provide unintended checkboxes

Settings[24] and consequences.”[26] preselection[25]

Bait and switch “You set out to do one thing, but a diferent, undesirable thing happens instead.”[3]

Coercion “Threatening or mandating the user’s compliance.”[27]

Technique

wording “The act of guilting the user into opting into something. The option preselected to decline is worded in such a way as to shame the user into compli- checkboxes, ance.” [ 3 ] popup, wording “Adverts that are disguised as other kinds of content or navigation, in linking order to get you to click on them.” [ 3 ] “Attracting the user’s attention away from their current task by ex- audio, ploiting perception, particularly preattentive processing.” [27] popup color, “[. . . ] lead users in a certain direction, making the process toward the nested path alternatives a long and arduous process[. . . ]”[24] “[. . . ] the service provider prevents the user from doing so by ei- absent informather—unnecessarily complicating the account deletion experience, or tion, nested path by not providing any account deletion option at all. ”[22] “[. . . ] any manipulation of the user interface that privileges specific ac- preselected tions over others, thereby confusing the user or limiting discoverability checkboxes, highof important action possibilities [. . . ]”[25] lighting, nested path

Interruption

“Interrupting the user’s task flow.” [27] popup

“Hiding desired information and interface elements.” [27]

Privacy Zuckering “You are tricked into publicly sharing more information about yourself mandatory field than you really intended to.” [3] entries, progress bar, wording

Scarcity

“ ‘Scarcity’ [. . . ] signal the limited availability or high demand of a highlighting, product, thus increasing its perceived value and desirability”[ 7 ] wording Social proof “informing users of others’ behavior and shopping with others”[ 7 ] apply a concrete technique to such a description. Further examples of deceptive patterns that we cannot map to a technique are: cuteness of robots, entrapping, two-faced and sneaking. We present the resulting overview of deceptive patterns that we can automatically detect in Table 1.

Attracting/ “A mechanism which tries to draw attention to certain options or color, fonts, high

Reducing atten- information with the use of highlighting.”[21] lighting, images, tion popup, wording

finan- “Choice architects can intervene with the perception of financial efort. preselected

Examples include postponing costs to the future without changing the checkboxes, actual final costs.”[21] wording

the “As users do not want to stick out, they have a tendency to do what highlighting, the majority of people does.”[ 21 ] wording Hiding informa- “Make undesirable options or information harder to see.”[ 21 ] tion “The act of making an attribute more salient (e.g., weight, price or color, fonts, highof color)”[ 21 ] lighting, images “Make incentives more salient or visible, so they are more efective and of prominent.”[ 21 ] color, fonts, highlighting, images time “When an option is presented as only being available for a certain highlighting, amount of time it is perceived as more important and scarce.”[ 21 ] wording “Describes the phenomena of users weighting losses higher than win- wording nings (e.g., the loss of €100 is worse than winning €100)”[ 21 ]

Make resources “Announcing limited availability of an option increases the probability highlighting,

scarce of users committing to choosing it.”[ 21 ] wording

Providing Feed- “Providing users with feedback when they are doing well or making popup, progress

back mistakes”[ 2 ] bar, wording

Remind of so- “Influence users by stating the social expectation in a specific situation highlighting,

cially desirable (e.g., need to vote)”[ 21 ] wording concepts

Setting defaults “Preselecting options by setting default options”[2]

preselected checkboxes

Suggesting alter- “Provide the users with alternatives that they might not have consid- highlighting,

natives ered at this point (e.g., cheaper camera with the same resolution)”[ 21 ] popup, wording Using visuals to “Through the use of optical illusions the perception and judgment color, fonts, highdeceit of options is altered, so they appear more salient than they actually lighting, images are.”[ 21 ] visuals “The salience of information is increased with the use of visual efects increase (e.g., colors, pictures, signs or fonts).”[ 21 ] color, fonts, highlighting, images Increase salience attribute Increase salience incentives Limited window Loss aversion Using to salience Visible goals Warning “Make simple performance measures clearly visible so that people can color, highlightimmediately assess their performance against a goal state (e.g., clearly ing, progress bar displaying manufacturing output and goals in factories is often, by itself, suficient to increase productivity).”[28] “Warn the users with the help of visuals or other means that emphasize highlighting, the problem at hand.”[ 21 ] popup, wording

Similarly, we proceed with Table 2: we search for empty cells in the column Technique, i.e., for nudges that are too broad in their meaning or could not be implemented with the techniques from Step 2. If we ifnd such a nudge, we delete the whole row. Examples for such nudges are: decrease vagueness and ambiguity, providing multiple viewpoints, simplification and understanding mapping. In Table 2, we show the remaining nudges. These are nudges that we can automatically identify.

Nudge

Attracting attention, Financial efort, Following the herd, Limited time,

Loss aversion, Scarce resources, Feedback, Socially desirable concepts,

Alternatives, Warning

Attracting attention, Following the herd, Hiding Information, Salience audio, color,

of attribute, Salience of incentives, Limited time, Scarce resources, fonts, highlight

Feedback, Socially desirable concepts, Alternatives, Visuals to deceit, ing, images,

Visuals to increase salience, Visible goals, Warning popup

Default Settings

Financial efort, Defaults

Bait and switch

Attracting attention, Feedback, Alternatives, Warning

Confirmshaming Attracting attention, Financial efort, Following the herd, Limited time, preselected

Loss aversion, Scarce resources, Feedback, Socially desirable concepts, checkboxes,

Defaults, Alternatives, Warning popup, wording

Step 4. Comparison In this step, we transfer the Deceptive Pattern and Technique columns from Table 1 to Table 3. Next, we identify any nudges in Table 2 that utilize the same techniques as the deceptive patterns. Upon finding such matches, we highlight the corresponding nudges in Table 2 in

Attracting attention, Hiding Information, Salience of attribute, Salience audio, of incentives, Feedback, Alternatives, Visuals to deceit, Visuals to in- popup crease salience, Visible goals, Warning color,

Attracting attention,Financial efort, Following the herd, Salience of preselected

attribute, Salience of incentives, Limited time, Scarce resources, Socially checkboxes, highdesirable concepts, Defaults, Alternatives, Visuals to deceit, Visuals to lighting, nested increase salience, Visible goals, Warning path

Attracting attention, Feedback, Alternatives, Warning popup

Attracting attention, Following the herd, Hiding Information, Salience audio, color,

of attribute, Salience of incentives, Limited time, Scarce resources, fonts, highlight

Feedback, Socially desirable concepts, Alternatives, Visuals to deceit, ing, images,

Visuals to increase salience, Visible goals, Warning popup

Attracting attention, Hiding Information, Salience of attribute, Salience color, of incentives, Visuals to deceit, Visuals to increase salience, Visible goals nested path fonts,

Attracting attention, Financial efort, Following the herd, Limited time, mandatory field

Loss aversion, Scarce resources, Feedback, Socially desirable concepts, entries, progress

Alternatives, Visible goals, Warning bar, wording

Attracting attention, Financial efort, Following the herd, Salience of highlighting,

attribute, Salience of incentives, Limited time, Loss aversion, Scarce wording resources, Feedback, Socially desirable concepts, Alternatives, Visuals to deceit, Visuals to increase salience, Visible goals, Warning

Attracting attention, Financial efort, Following the herd, Salience of highlighting, attribute, Salience of incentives, Limited time, Loss aversion, Scarce wording resources, Feedback, Socially desirable concepts, Alternatives, Visuals to deceit, Visuals to increase salience, Visible goals, Warning

bold. Then, we put the names of these nudges in the Nudge column for the relevant row in Table 3, displaying the findings of our analysis.

Step 5. Identifying Diferences Having Table 3, we begin to evaluate if we can distinguish nudges from deceptive patterns. If there are some unmarked nudges available in Table 2, we can identify these nudges as distinguishable from deceptive patterns by the techniques they use. However, all nudges are marked in bold, i.e., there is no nudge that we can distinguish from deceptive patterns. In contrast, there are deceptive patterns that do not share techniques with any nudge from Table 2. We identify such deceptive patterns by empty cells in the column Nudge. These deceptive patterns are: absent information, audio, nested path, mandatory field entries. Therefore, based on this information, we conclude that it is not technically feasible to distinguish nudges from deceptive patterns. This is because there are no techniques in nudges which are not also used by deceptive patterns. Thus, a crawler cannot discern whether a technique is used by a nudge or by a deceptive pattern.

7. Related Work

There are several papers that focus on the study of deceptive patterns or nudges. There are also some papers that address the fact that there is overlap between deceptive pattern and nudges. However, to the very best of our knowledge, there is no work that compares deceptive patterns and nudges systematically and from a technical perspective. In the following, we discuss only those papers that at least mention the diferences and similarities between the two concepts.

Narayanan et al. [29] state that deceptive patterns are a result from three decade-long trends: retail (deceptive practices), research and public policy (nudging), and the design community (growth hacking). The authors classify research on nudges as a source of deceptive patterns. Here, nudging refers to the 1970s, when “the heuristics and biases literature of behavioral economics sought to understand irrational decisions and behaviors” [29]. Thus, the authors see a connection between deceptive patterns and nudges, based on the idea that deceptive patterns have evolved from, among other things, nudging. However, they focus only on a few examples and a few designs, and do not go beyond that.

Morrison et al. [30] highlight that dark patterns intersect deceptive techniques, nudging, and social engineering, building on Narayanan et al. [29]. In contrast to our work, they refer to mental models and do not go beyond that point either.

In other papers, the authors do not directly link deceptive patterns and nudges, but do acknowledge their presence. For example, Weinzierl [ 6 ] introduces “dark nudging” as a term to bridge deceptive patterns with the behavioral psychology principles underlying nudging. In [ 5 ], the authors discuss the blurred lines between persuasion and manipulation, noting the possibility that designers might unknowingly create interfaces with deceptive patterns, highlighting the challenges in diferentiating between deceptive patterns and nudges. Jesse and Jannach [ 21 ] remark on the dificulty of clearly separating nudging from persuasion, emphasizing that both are aimed at influencing user behavior and decisions. The key distinction they note is whether the influence is beneficial or detrimental to the user.

8. Conclusion

We investigated the technical possibility of diferentiating nudges from deceptive patterns in web analysis. Our developed methodology showed that, technically, both concepts are indistinguishable, suggesting crawlers may capture both deceptive and non-deceptive patterns alike.

In future work, we will examine what additional features or their combination we can integrate to enable the diferentiation. For this, we will first analyze whether and how we can use wording to automatically distinguish between deceptive patterns and nudges.

Computers in Human Behavior Reports 3 (2021). doi:10.1016/j.chbr.2020.100052. [22] C. Bösch, B. Erb, F. Kargl, H. Kopp, S. Pfattheicher, Tales from the dark side: Privacy dark strategies and privacy dark patterns., Proceedings on Privacy Enhancing Technologies 2016 (2016) 237–254. doi:10.1515/popets-2016-0038. [23] S. Greenberg, S. Boring, J. Vermeulen, J. Dostal, Dark patterns in proxemic interactions: a critical perspective, in: Proceedings of the 2014 Conference on Designing Interactive Systems, 2014, pp. 523–532. doi:10.1145/2598510.2598541. [24] N. C. Council, Deceived by design, how tech companies use dark patterns to discourage us from exercising our rights to privacy, Norwegian Consumer Council Report (2018). https://fil. forbrukerradet.no/wp-content/uploads/2018/06/2018-06-27-deceived-by-design-final.pdf. [25] C. M. Gray, Y. Kou, B. Battles, J. Hoggatt, A. L. Toombs, The dark (patterns) side of ux design, in: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, 2018, pp. 1–14. doi:10.1145/3173574.3174108. [26] C. M. Gray, S. S. Chivukula, A. Lee, What kind of work do" asshole designers" create? describing properties of ethical concern on reddit, in: Proceedings of the 2020 ACM Designing Interactive Systems Conference, 2020, pp. 61–73. doi:10.1145/3357236.3395486. [27] G. Conti, E. Sobiesk, Malicious interface design: Exploiting the user, in: Proceedings of the 19th International Conference on World Wide Web, 2010, pp. 271–280. doi:10.1145/1772690. 1772719. [28] W. Lidwell, K. Holden, J. Butler, Universal principles of design, revised and updated: 125 ways to enhance usability, influence perception, increase appeal, make better design decisions, and teach through design, Rockport Pub, 2010. [29] A. Narayanan, A. Mathur, M. Chetty, M. Kshirsagar, Dark patterns: Past, present, and future: The evolution of tricky user interfaces, Queue 18 (2020) 67–92. doi:10.1145/3400899.3400901. [30] B. Morrison, C. Sengul, M. Springett, J. Taylor, K. Renaud, Mental models of dark patterns (2021). https://spritehub.org/wp-content/uploads/2021/12/SPRITE_Lit_Review8.pdf.

References

[1] S. Englehardt , A. Narayanan , Online tracking: A 1-million-site measurement and analysis , in: Proceedings of ACM CCS 2016 , 2016 , pp. 1388 - 1401 . doi: 10 .1145/2976749.2978313.

[2] M. Weinmann , C. Schneider , J. v. Brocke , Digital nudging, Business & Information Systems Engineering 58 ( 2016 ) 433 - 436 . doi: 10 .1007/s12599-016-0453-1.

[3] H. Brignull , Deceptive patterns, 2023 . www.deceptive.design.

[4] M. Martini , C. Drews , P. Seeliger , Q. Weinzierl , Zeitschrift für digitalisierung und recht ( 2022 ).

[5] J. Luguri , L. J. Strahilevitz , Shining a light on dark patterns , Journal of Legal Analysis 13 ( 2021 ) 43 - 109 . doi: 10 .1093/jla/laaa006.

[6] Q. Weinzierl , Neue Zeitschrift für Verwaltungsrecht-Extra ( 2020 ).

[7] A. Mathur , G. Acar, M. J. Friedman , E. Lucherini , J. Mayer , M. Chetty , A. Narayanan , Dark patterns at scale: Findings from a crawl of 11k shopping websites , Proceedings of the ACM on Human-Computer Interaction 3 ( 2019 ) 1 - 32 . doi: 10 .1145/3359183.

[8] K. Bongard-Blanchy , A. Rossi , S. Rivas , S. Doublet , V. Koenig , G. Lenzini, ” i am definitely manipulated, even when i am aware of it. it's ridiculous!”-dark patterns from the end-user perspective , in: Designing Interactive Systems Conference 2021 , 2021 , pp. 763 - 776 . doi: 10 .1145/3461778. 3462086.

[9] F. J. Costello , J. Yun , K. C. Lee , Digital dark nudge: An exploration of when digital nudges unethically depart , in: Proceedings of the 55th Hawaii International Conference on System Sciences , 2022 . doi:10125/79868.

[10] R. H. Thaler , C. R. Sunstein , Nudge, PENGUIN BOOKS , 2021 , pp. 6 - 8 .

[11] C. Meske , T. Potthof, The dinu-model-a process model for the design of nudges , Proceedings of the 25th European Conference on Information Systems (ECIS) ( 2017 ).

[12] P. Graßl , H. Schrafenberger , F. Zuiderveen Borgesius , M. Buijzen , Dark and bright patterns in cookie consent requests , Journal of Digital Social Research 3 ( 2021 ) 1 - 38 . URL: https://jdsr.se/ojs/ index.php/jdsr/article/view/54. doi: 10 .33621/jdsr.v3i1. 54 .

[13] S. Schomberg , T. J. Barev , A. Janson , F. Hupfeld , Ansatz zur umsetzung von datenschutz nach der dsgvo im arbeitsumfeld: Datenschutz durch nudging , Datenschutz und Datensicherheit-DuD 43 ( 2019 ) 774 - 780 . doi: 10 .1007/s11623-019-1204-5.

[14] watson.de, Green Nudging: Schon kleine Anreize bringen Menschen zum Energiesparen , 2022 . https://www.watson.de/!168653241? utm_medium=social-user&utm_source=social_app , urldate = 2023 -03-06.

[15] A. Caraban , E. Karapanos , D. Gonçalves , P. Campos, 23 ways to nudge: A review of technologymediated nudging in human-computer interaction , in: Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems , 2019 , pp. 1 - 15 . doi: 10 .1145/3290605.3300733.

[16] D. Djurica , K. Figl , The efect of digital nudging techniques on customers' product choice and attitudes towards e-commerce sites , Americas Conference on Information Systems 2017 ( 2017 ) 1 - 7 .

[17] A. Kozyreva , S. Lewandowsky , R. Hertwig , Citizens versus the internet: Confronting digital challenges with cognitive tools , Psychological Science in the Public Interest 21 ( 2020 ) 103 - 156 . doi: 10 .1177/1529100620946707.

[18] A. Acquisti , I. Adjerid , R. Balebako , L. Brandimarte , L. F. Cranor , S. Komanduri , P. G. Leon , N. Sadeh , F. Schaub , M. Sleeper , et al., Nudges for privacy and security: Understanding and assisting users' choices online , ACM Computing Surveys (CSUR) 50 ( 2017 ) 1 - 41 . doi: 10 .1145/3054926.

[19] A. Mathur , M. Kshirsagar , J. Mayer , What makes a dark pattern... dark? design attributes, normative considerations, and measurement methods , in: Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems , 2021 , pp. 1 - 18 . doi: 10 .1145/3411764.3445610.

[20] N. C. on Informatics , L. ( CNIL), Shaping choices in the digital world, 2020 . https://linc.cnil.fr/sites/ default/files/atoms/files/cnil_ip_report_06_shaping_choices_in_the_digital_world.pdf, urldate = 2023 -03-06.

[21] M. Jesse , D. Jannach , Digital nudging with recommender systems: Survey and future directions,