Abstract
European General Data Protection Regulation requires organisations to request the data subject’s consent for personal data processing. Data controllers must be able to demonstrate valid consent was obtained (‘transparency’). Media often struggle to meet GDPR requirements in practice. We identified several issues with existing consent procedures amongst which a need for trustworthy approaches to record and track consent. In this article, we evaluate a specific transparency initiative: a Personal Data Receipt (PDR) for news personalisation. We investigated how European media users and media professionals evaluated the PDR. We conducted qualitative surveys and interviews to explore and describe individuals’ viewpoints on/responses to the PDR. The main strengths highlighted in this study are: GDPR compliance and improved data processing transparency which leads to more control and user trust. PDR weaknesses are mainly related to users not reading the receipt, lack/overload of information, and design issues. Based on our findings, we identified missing elements and formulated recommendations for PDR improvement to optimise consent strategies. By examining how individuals responded to this specific transparency tool, and rhetorical tactics connected to it (placation, diversion, jargon, and misnaming), our study provides informed suggestions for ways out of digital resignation (Draper & Turow, 2019).
Keywords
Introduction
A standard question after money-based transactions is: Would you like a receipt? This records the price paid/received for products or services. Receipts provide “bedrock for trust, autonomy and freedom from abuse” in consumer law (Lizar & Hodder, 2014, p. 648). This is similar for online purchases. But what about companies offering ‘free’, (personalised) media content in exchange for disclosing personal data (age, gender, interests)? Exchanging services for personal data is a practice currently applied by nearly all media organisations. European General Data Protection Regulation (GDPR) requires organisations to request consent for personal data processing to protect the consumer. Nevertheless, we identified several issues with consent procedures. Most users agree with privacy policies without reading them (Lizar & Wunderlich, 2014; Meadows & Hatzakis, 2018; Nati, 2018; Vanhaelewyn & De Marez, 2018). The format of consent requests is often problematic. Terms and conditions (T&Cs) are generally unclear, cumbersome to read, and hinder information on altering choices (Lizar & Hodder, 2014; Lizar & Pandit, 2019; Nati, 2018; Santolalla, 2018). Digital apps and services’ consent procedures are “structurally and practically flawed” (LeVasseur & Maler, 2019, p. 52). Consent also often remains unrecorded and untraceable for users (Lizar & Hodder, 2014; Lizar & Pandit, 2019). Organisations obtain ‘green light’ for using consumer data, yet users have limited control and do not understand/remember to whom they gave consent and for what purposes (Meadows & Hatzakis, 2018; Santolalla, 2018; Su et al., 2016). Users are also unaware of personal data value (Van Buggenhout & Van den Broeck, 2020). Furthermore, personal data has economic value for media companies, but public awareness is low about societal benefits like economic growth (Styliari & Nati, 2016; Su et al., 2016). This results in an unfair value-exchange between industry and users (Van Buggenhout & Van den Broeck, 2020). Control is concentrated in a few hands – ‘siloed’ (Hardjono & Pentland, 2019). Organisations exploit user data for profiling, product/service development (personalisation), marketing, customer relationship management, and user research (Ibid; Nati et al., 2019; Politou et al., 2018; Van Buggenhout & Van den Broeck, 2020). Users on the other hand value privacy and desire control of personal data processing by organisations and governments, but they feel unable to do so (Draper & Turow, 2019). “Surveillance is inescapable”, this sense of futility or helplessness is conceptualized as ‘digital resignation’ (Ibid, p. 1825).
Lack of digital transparency, security, and accountability are critical challenges for personal data processing trustworthiness (Lizar & Pandit, 2019). Organisations must prevent misuse and increase user trust (Nati, 2018; Nati et al., 2019; Patachi, 2018). They should improve transparency and control for users (Nati, 2018). Organisations need best practices for communicating about personal data processing value, benefits, and risks towards users (Van Buggenhout, Van den Broeck & Ballon, 2020; Van Zeeland et al., 2019).
In this article, we evaluate a specific transparency initiative: a Personal Data Receipt (PDR) for news personalisation. This could be considered a viable solution for organisations to improve user trust in personal data processing, like proof-of-purchase receipts (Styliari & Nati, 2016). PDR track information being collected, linked to processing purposes and record consent (Meadows & Hatzakis, 2018; Lizar & Wunderlich, 2014). This eliminates the problem of (not) reading T&Cs for users, being unaware of information you shared, and reconstructing the data trails you left behind (Lizar & Pandit, 2019; Styliari & Nati, 2016). Organisations’ adoption and PDR implementation are low, however, like user awareness of this tool (Leu, 2019; Santolalla, 2018). These insights fuelled two research questions:
Do PDR provide actual transparency for media users? Do professionals assess PDR as effective transparency tool to improve user trust in personal data processing?
Our study focused on the PDR applied in the European H2020 project Content Personalisation Network (CPN). The PDR is a standardised, digital/human-readable transparency tool to improve user control (Nati, 2018). Effective to communicate “why personal information is gathered and how it is used” (Ibid, pp. 8–11). CPN tested a personalised news application (2017–2020). “A new, trustworthy approach to personalise digital content” (
“a record of the permissions they gave the CPN platform to hold and process their data […] Any time the user changes their permissions within the platform, they will receive a new PDR with their updated consent and preferences, so they can check how these have changed over time” (ibid) (Fig. 1).
PDR (CPN Consortium, 2019).
Our findings result from mixed methods research (inductive). We examined European media users’ and professionals’ responses to PDR (qualitative surveys and interviews), and the connected rhetorical tactics (Draper & Turow, 2019). We identified PDR strengths and weaknesses along with recommendations for its improvement. Our study contributes to consent standard refinement, user and institutional data literacy, and the establishment of collaborations using PDR as prerequisite in data transactions (Nati, 2018; Styliari & Nati, 2016). We provide informed suggestions for best practices in transparency initiatives and ways out of digital resignation (Draper & Turow, 2019).
After this introduction, we delve into the conceptual framework of the study. Based on literature, we substantiate importance of personal data processing transparency, review (dis)advantages as well as similarities and differences between consent receipts and PDR. Next, we describe the methods and report the results of media users’ and professionals’ appreciation of PDR. We integrate and discuss the meaning of our findings in Section 5, followed by a conclusion.
We follow GDPR art. 4 (1), (2), (4), (7), (8), and (11) definitions of the following concepts, to evaluate personal data processing transparency. Applied to the media context we are focusing on in this article, processing is personal data collection, storage, alteration, use, disclosure, etc. In the case of PDR, controllers/processors are media organisations which determine processing means and purposes (i.e., profiling), and process user data (‘subjects’). Users’ consent for personal data processing must be freely given, specific, informed, and unambiguous. In ethical and moral-philosophical sense, consent is necessary to permit conduct that would otherwise be ‘wrongful’ or ‘impermissible’ (Miller & Wertheimer, 2010; Simmons, 1998). Media must process data in ways that users reasonably expect and demonstrate GDPR compliance (GDPR Recital 74; ICO, s.d.). It should be transparent to users what personal data are collected, to what extent information is or will be processed, by whom, and for what purposes (GDPR Recital 39, 42). “Transparency empowers data subjects to hold data controllers and processors accountable and exercise control over their personal data” i.e., provide/revoke consent (WP29, 2018, p. 5). Media must provide clear, concise, comprehensible (plain language), and easily accessible information to users (GDPR art. 12–14; WP29, 2018; Nati et al., 2019). Efficient and succinct communication avoids ‘information fatigue’, and users prefer immediate access rather than having to scroll through long T&Cs (WP29, 2018). Media must take appropriate measures to inform users, GDPR does not prescribe the format or modality – ‘how’ to provide clear information (GDPR art. 12 (1); WP29, 2018).
These concepts provide clear grounds and define the playing field for media to process European users’ data. Several organisations however struggle to apply GDPR in practice. Examples of efficient communication are lacking (Van Buggenhout et al., 2020; Van Zeeland et al., 2019). Communication should be open, clear, and without deception (Butarbutar, 2020). Transparency and control positively influence users’ willingness-to-share information, this enhances organisations’ access to quality user data (Ibid; Nati, 2018; Robinson, 2018). Some organisations introduced transparency initiatives for users to view and amend the information companies collect i.e., Google privacy dashboard and Facebook ad preferences (Draper & Turow, 2019). These tools often (are intended to) obfuscate data practices; cultivate digital resignation, rather than provide effective transparency (Ibid). Four rhetorical tactics characterize the obfuscation process (Ibid):
Placation: Falsely appease concerns, suggest user empowerment and control, while giving little insight into actual practices. Diversion: Shift individuals’ focus away from controversial practices, not providing the necessary information to achieve actual transparency. Jargon: Complex terminology to discourage users from engaging with information about data processing. Misnaming: Misleading labels to impede insight in industrial practices.
There is a clear need for privacy-friendly business models to generate economic value from data processing (Meadows & Hatzakis, 2018). “Web standards have emerged for the specification and implementation of consent procedures in online environments” (Politou, Alepis & Patsakis, 2018, p. 8). Data protection regulations and consent standards being GDPR, User Managed Access (UMA) and Minimum Viable Consent Record (MVCR)1
Kantara Initiative developed UMA1.0 and MVCR0.7.1 in 2015, and UMA2.0 in 2018 (Hardjono et al., 2015; Lizar, 2016; Maler et al., 2018).
Kantara Initiative created the Consent Receipt Specification (CR1.0) in 2017 (Greig, 2017; Lizar & Turner, 2017; Politou et al., 2018; Lizar & Wunderlich, 2014; Styliari & Nati, 2016). No effective standard/common format existed for recording consent nor for providing consent receipts; users could not track consent, monitor data processing, or hold controllers accountable i.e., for privacy breach (Greig, 2017; Lizar & Turner, 2017). CR1.0 provided a standardised consent notice when users agreed to personal data processing – like “a cash register receipt as a personal record of a purchase transaction” (Greig, 2017). The ‘artefact’ describes to users what information controllers collect, and for what purposes (Hardjono & Maler, 2017). CR have three content fields: transaction details (consent timestamp, receipt ID), transaction parties (controllers’ and users’ contact information), and processing details i.e., data categories, processing purposes and third-party disclosure (Ibid). Consent receipts are personal data, it “links multiple data sources with an identifier” (Ibid, p. 18). CR1.0 was a “promising standard” but not GDPR compliant considering terminology and human readability (Santolalla, 2018). Table 1 summarises CR1.0 strengths and weaknesses mentioned in previous studies. The overview illustrates consent receipts have several benefits, but there are still opportunities for improvement.
CR1.0 evaluation
DigiCat developed PDR inspired by CR1.0 (Nati, 2018; Politou et al., 2018; Styliari & Nati, 2016). The developers leveraged the existing standard to “generate awareness of a consumer-centric consent process for increasing consumers’ trust in organisations” (Styliari & Nati, 2016, p. 4). DigiCat describes PDR as:
“a human-readable digital record summarising in a simple and clear way what personal data an organisation is collecting about an individual, for what purpose, how it’s stored, for how long and if any third-party sharing is allowed” (Nati, 2018, p. 5).
PDR are provided before or after consent is given. PDR aim to improve data processing transparency, by clearly informing users. Simultaneously, PDR should be less overloading for end-users than T&Cs for example, as they offer more transparency through less text.
PDR concept viability was tested at DigiCat (Nati, 2018; Styliari & Nati, 2016). User-feedback demonstrated PDR should contain which data is collected, processing purposes, third-parties, retention periods, and contact details for users to request data removal or revoke consent (Ibid). “The main benefit is perhaps societal, by triggering a change in the way we conduct data transactions” (Styliari & Nati, 2016, p. 15). PDR comply with GDPR requirements and terminology (Nati, 2018). PDR are machine and human-readable, “issued as mobile-friendly email when customers join a new digital service” – “ideal format” (Jesus, 2020, p. 25385; Nati, 2018, p. 5). PDR include visual and textual information like icons, which are “meaningful and easy-to-understand” (Nati, 2018; Styliari & Nati, 2016, p. 4). Addressing “consumer needs for simple privacy statements” (Nati, 2018, p. 1). PDR enhance data literacy, user empowerment, control, and “educate individuals […] to take action when they are not satisfied about the use of it” (Ibid, pp. 10–11). DigiCat’s PDR and Kantara’s consent receipt benefits are similar, but PDR overcome CR1.0 weaknesses. PDR are GDPR compliant and machine/human readable.
Methods
We evaluated the PDR as implemented in the European H2020 project CPN (real-life setting). We focused on media users’ and professionals’ appraisal of PDR. Actual user data was collected for a new digital service (personalised news). Users were informed via a customized PDR, after consent to data processing by CPN (Fig. 1). This article presents an in-depth view on how individuals responded to this specific transparency initiative. The participants had no prior knowledge, nor experience with PDR.
Exploring end users’ perspective
We surveyed CPN app testers (Feb–Mar 2020). The news app provided personalised news in English, Greek, and Dutch. CPN partners accommodated news in three countries: Deutsche Welle (DW – English content) in Germany, DIAS in Cyprus – Greek content, and VRT in Belgium – Dutch content. The survey was conducted in the three languages and examined how users experienced personalised news. We included an item (Net Promoter Score – ‘NPS’) to determine PDR transparency2
Exclusively for DW and DIAS testers (mobile and web CPN-app). VRT MYNWS users did not receive PDR.
As second end-user activity, five students in communications and young media professionals evaluated the PDR during an online brainstorm, organised for the [imec-SMIT-VUB] Media Economics Postgraduate 2020. Participants formulated feedback from a consumer perspective via polling and group discussion (in real-time). We recorded audio/video (anonymised ranscripts). We manually coded, clustered and analysed research data following the grounded theory principles of open, axial, and selective coding (Glaser & Strauss, 2017).
To examine media professionals’ perspective, we surveyed an expert panel4
Ten Flemish media professionals, eight academics, one media advisory board, and Belgian data protection regulator.
Three VRT interviewees, eight DW respondents, and five DIAS interviewees.
PDR NPS-scores (end users)
*Did not explain their answer.
PDR evaluation (end users)
PDR appreciation by end users
Table 2 summarises the surveys6
We descriptively analysed and reported the survey data (CPN-/web-app). There were too few survey responses to perform in-depth statistical analysis.
“Users cannot be sure that only the above personal data is processed. However, it is positive that users have a clear description of how long data will be processed in an electronic document.” (DIAS CPN-app tester).
Our results indicate mixed user perspectives on the role of PDR regarding personal data processing transparency. Arguments in favour are the icons and bullet points make it simple and user-friendly. PDR are more comprehensive than other approaches for providing information about personal data processing. For example, T&Cs contain legal text that most people do not understand. PDR are useful for people who are digitally illiterate, but with a more extensive explanation. Arguments against are i.e., it is still “vague”. A participant suggested information is missing about the monetary value of personal data:
“Concrete numbers with maybe a sum in dollars or euros […] It only says what is stored but not what the value of it all is to a media organisation or to yourself”.
This viewpoint raised interest and questions from the group like “How can you quantify that?” and “Isn’t it difficult to put a price on someone’s data? That will be different from person to person”. The participant replied that quantifying personal data value is not difficult i.e., “by comparing ad income divided by [the number of] users”. Personal data prices should be calculated per individual. PDR should be ‘personalised’ with prices adapted to individual users: “data is personal so why can’t the form be personal then?”. Participants unanimously approved this idea. We asked how they evaluated the approach of emailing PDR when users register for a service and if settings change. A participant responded PDR enhance data literacy, “you are aware of everything that is going on in your account”.
PDR NPS-scores (expert panel)
PDR evaluation (expert panel)
Table 4 summarises survey and PDR-question response count, NPS-categories, and PDR NPS-score. We discerned themes and bundled quotes of promoters, passives, and detractors (Table 5). Most promoters described PDR strengths related to transparency, trust, consent, control, accountability, and usefulness for users. Two promoters identified PDR weaknesses. PDR comprehensiveness depends on users’ data literacy. More information is needed, but (not) reading PDR is problematic. “It can be done in more detail, but nobody will read that” (data protection and privacy expert at public broadcaster). Passives illustrated PDR strengths and weaknesses; detractors put forward mostly weaknesses. Five of twenty respondents across different NPS-categories expect PDR improvements and future iterations (three detractors, one passive and one promoter). They indicated i.e., ‘good start’, ‘test project’, ‘interesting concept’ and ‘step in the right direction’.
From a professional perspective, the sixteen interviewees agreed media should inform users when they create a profile about them. This is mandatory under European rules. The media professionals think people should have control over this. It is better if users create their own profile and choose (not) to share data with certain organisations or for specific purposes. ‘Opt-in’ is preferred. Maybe users are already aware:
“We believe that once users visit a webpage and sign in, they know that a profile is being created for them. But it is proper that there is some informative material that they are entering a ‘personalization phase’ (DIAS interviewee)”.
Whether users should be informed depends on “what personal data is collected and how anonymous this is” (VRT journalist). Media grudgingly provide information about this, “it is unethical for a news organization to do that without the user being aware of it” (DIAS interviewee). DW respondents commented user profiles must be ‘transparent’, ‘accessible’, ‘adjustable’, and users should be able to ‘delete’ their profiles. It is important to inform users in a general way, communicate about how the profile was created, what it entails, and explain how data is analysed. There should be transparency towards users, from an ethical perspective, this is expected by audiences or required by law. Media should inform their users about personalisation criteria used to recommend news content. A DW respondent partially agreed with this statement and commented “if the user wants to know more (s)he should be able to request and receive the information easily”. Transparency is important, “otherwise it [personalisation] will be criticised quickly” (VRT journalist). Is it unethical to do personalisation under the radar? “That [transparency] is quite a discussion and depends a bit on how far you go in it” (VRT data analytics expert). A DW interviewee recommended to stop personal data processing solely based on selling more products and services. Van Buggenhout and Van den Broeck (2020) encountered this practice in their research on personal data value. Media organisations cast a ‘wide net’ and collect more data than needed to provide personalised services (Viala, 2018).
When it comes to user notifications, interviewees described standard, contrived solutions such as email or pop-ups. For example, a confirmation email when users register, explaining which data is stored about you. It can be a blog post, or an email automatically sent after users ‘opted-in’ to personalised news. Notwithstanding, not every user should be informed in a personal way. This information can be displayed on the tab/page where personalised news is provided. DIAS interviewees furthermore proposed “a link with continuous notifications” or “notification at the time the user visits the organisations’ website” (continuous vs one-off notifications). Explanations should be easy to understand. “Not informing is a worse thing than informing” (VRT social media and video production expert). In the case of CPN, two media professionals disagreed users should be informed about which personal data is used for news personalisation:
“I do not think it is necessary, though it might be a good practice. I see it from the opposite viewpoint, not that of the user. […] From the viewpoint of users, it would be good. Maybe there should be a list with news categories from which they can choose and give their consent” (DIAS interviewee).
There is a common sense among the experts that users should be informed, but the specific format in which this could or should happen differs. We confronted them with PDR as a specific transparency tool to inform users about personal data processing. All interviewees positively evaluated PDR. For example, it provides overview to users. “Something like this is often missing from data handlers or it is long and complex” (DW professional). Interviewees affirmed PDR improve transparency and trust. For example, “a good approach to ensure mutual trust” (DIAS interviewee), “boost confidence” (VRT data analytics expert), and “open” (VRT social media and video production expert). A DIAS interviewee mentioned PDR increase accountability, “like having a lawyer signing papers”, for (consenting to) personal data processing. Correspondingly, a VRT interviewee asked: “I assume that something like that evolves or expands as more data is collected over time?”. Users indeed receive a revised PDR (email) if data processing settings change. PDR user-friendliness decreases “if it changes too often and you get too much information” (VRT interviewee). Two interviewees indicated the issue of users (not) reading the PDR. “It’s yet another email” (DW interviewee). The PDR terminology is often legal text and discouraging for people to read. There are missing elements, PDR are “unspecific in regard to what was actually stored (no links to repositories/overviews) and what exactly it is being used for” (DW interviewee). There are no technical barriers for PDR implementation from media professionals’ perspective, but the benefits, complexity and obstacles for users need further investigation before media are willing to actualize PDR provision. “Take the current cookie messages you get on websites if you are accessing them from Europe. People get annoyed by those if they are not easy to handle” (DW interviewee).
Discussion
We investigated how media users and media professionals evaluate PDR considering transparency and trust. Our guiding research questions were: do PDR provide actual transparency from media users’ perspective, and do professionals expect PDR to effectively improve user trust in personal data processing? In this study, we found that PDR is appreciated as a valuable transparency tool by media users and professionals (clear, comprehensive, and trustworthy), a step in the right direction, but PDR improvement is also needed i.e., regarding standardization, provision, usability, and user experience. Based on the qualitative survey findings NPS-scores are mixed (ranging from low to medium,
The main
PDR help media to achieve GDPR compliance and improve privacy. Personal data processing transparency and user trust increases. PDR value depends on the user’s trust in the controller (Styliari & Nati, 2016). Users perceive PDR created by trustworthy organisations as more valuable than PDR provided by untrustworthy organisations, if provided at all. Standardisation is important. PDR should be “universal” (Styliari & Nati, 2016, p. 13). Consent and control for users, accountability for controllers. Convincing approach if users’ attitude is positive towards data processing. PDR provision creates a barrier for personal data misuse. Informativeness and readability (use of icons). PDR are more comprehensive than other approaches (T&Cs). Also demonstrated by Nati (2018) and Styliari & Nati (2016). Clear description of data processing benefits for users. Improves data literacy. Useful to exercise data rights. Simple, user-friendly format and lay-out.
Study participants put forward
Not reading PDR. Most CPN users did not remember having received PDR (Wagner et al., 2019). PDR testing at DigiCat demonstrated 50% of individuals who received PDR ( Unconvincing approach if users’ attitude is negative towards data processing and/or the controller itself. Fear of personal data misuse. More information is needed regarding PDR implementation. Consumers still question if media are genuinely transparent and provide the necessary information i.e., resale to third parties. Language (in English only). Too technical, general, and vague terminology. Meaningful PDR are clear, simple, and unambiguous (Jesus, 2020; Nati, 2018). PDR should be short and include web links in specific sections with detailed information (Styliari & Nati, 2016). Impersonal format and layout. User-friendliness decreases if people receive PDR when settings change (information overload). This corroborates Cooper’s (2017) statement “consent receipts are all over the place”, users need trustworthy storage solutions to consolidate all PDR into a single view. For example, PDR wallets or blockchain-based notice registries.
Is the PDR an effective transparency initiative or obfuscatory communication practice to cultivate digital resignation? We performed a post hoc analysis, but it could also be useful before the interviews. We aimed to interpret research data, explore, and describe participants’ perspectives, with a minimum of bias (‘not-knowing stance’) and maximum of empathy. Grounded in the research data, we understand PDR do not falsely appease user concerns, but more insight into actual data practices is necessary. For example, how data circulate in the broader systems that give them meaning (Draper & Turow, 2019). PDR do not divert individuals and are an effective tool to improve personal data processing transparency. Additional explanation is needed for users to understand the information to which they have been granted access (Ibid). The PDR terminology is too complex (jargon). PDR labels are not misleading.
PDR improvement
Based on our analysis, we identified missing elements and formulated recommendations for PDR improvement (Table 6). We advocate standardization of PDR format and lay-out. PDR should inform users on third parties’ access to user data for purposes like targeted advertising. PDR could also inform users about monetisation (data as ‘currency’). Acknowledging users’ right to be informed about personal data value could improve transparency (Malgieri & Custers, 2018). PDR could clarify the link between collected user data and generated insights. Finally, PDR should provide complete information on users’ data rights.
Our study contributes to privacy paradox literature. There are few viable solutions to span the gap between user intention and behaviour (Barth & de Jong, 2017). Users are concerned about privacy, but act to the contrary, agreeing with T&Cs without reading them (Ibid). PDR provide more transparency by offering relevant information more concisely. PDR can create privacy awareness and help users to avoid paradoxical behaviour (Pötzsch, 2009). The tool reminds users about their privacy intentions and supports people to make informed decisions about information disclosure (Ibid). PDR can be considered as Transparency Enhancing Technology (TETs) as it helps “a user to regain control over her personal sphere” (Hedbom, 2008, p. 67). PDR allow users and professionals to keep track on which information is stored, by whom it is used and for what purposes (auditing). PDR are an interactive tool; users can actively influence the stored data and processing (exercise data rights).
Despite their potential, PDR are not yet an ubiquitous transparency tool. Several challenges remain for widespread implementation by media and user adoption. Some organisations hesitate or do not implement PDR, as they reveal (too) much information to users with “one quick read” (Styliari & Nati, 2016, p. 15). “Even a brilliant and popular technical solution can be defeated by lousy regulations or a lack of international consensus” (Santolalla, 2018). After all, it is difficult to cooperate at national level. “One-off approaches rarely result in broad social change […] individual responses seldom succeed in undermining powerful systems” (Draper & Turow, 2019, p. 1833). PDR do not work as “stand-alone” (Nati, 2018, p. 6). “But alongside many consent receipts, used in every data transaction, they will be powerful and can change the way we trust organisations and share our data” (Styliari & Nati, 2016, p. 10). This creates an accountable and auditable ‘Web-of-Receipts’/WoR for users, and media can demonstrate that valid consent was obtained for personal data processing (Jesus, 2020). Providing digital receipts – ‘proof’ – each time users consent to personal data processing, creates an ‘audit trail’ for subjects and controllers (Morrow, 2017; Patachi, 2018). “Each will have undisputable evidence of the (personal data) transaction details” – like ‘normal’ purchases (Jesus, 2020, p. 25384). It improves digital transparency and trust (Styliari & Nati, 2016). This makes people more amenable to share information with media (Robinson, 2018). WoR will impact controllers’ consent procedures and reverse “the power relationship from solely organisations to a shared, “one on one”, responsibility” (Ibid, p. 25387). Transparency tools like PDR have business and technical trust implications (Hardjono & Maler, 2017). PDR usage and management is technically challenging and requires development, evaluation, and adaption of trustworthy storage solutions. Users should self-manage (store/retrieve) PDR in a ‘Receipt Wallet’ i.e., cloud storage or a trusted third party can aid users in managing the large number of receipts (Jesus, 2020). For example, the Consentua Consent Repository and the Open Notice Registry could be used for this (Cooper, 2017; Lizar & Hodder, 2014).
Based on our analysis, we see an important future role for PDR along with continuation of PDR research. If media implement the provided recommendations and users recognize PDR value, a top-down approach is likely needed to establish PDR adoption. Future studies should leverage media professionals and experts to expand PDR research and focus on current barriers for PDR adoption. For example, examining whether PDR provide too much transparency that can impede market adoption. Consent interactions are increasingly automated, categorizing consent considering scenario and outcome, PDR function as “bridge to bring all the receipts together into one place (a virtual place) […] dynamic store of consent” (Cooper, 2017). It would be fruitful to further investigate to which extent sector-wide support exists in the media for implementing PDR on regional, national, or European scale. Our research focused on media, but PDR would benefit all sectors that use personal data for personalisation and other purposes (banking, energy, health, etc). The study raises important questions for policymakers. For example, should PDR be legally required? It would be an important step towards transparency, but it also requires scrutinizing of existing PDR-storage solutions, which we see as an interesting topic for future work.
Scrutinizing the role of PDR in a broader context and considering societal debates, it is an improvement in terms of transparency and data literacy. On the other hand, the question of responsibility – should this be in the hands of users or companies? – remains. We argue that it is a shared responsibility. Users should not resign to a privacy paradox and must improve their understanding of data processing consequences. Companies must provide clear information about processing purposes. PDR provision nevertheless might result in a false sense of control (individuals can only agree to what the company proposes), and therefore also of user agency.
Trying to achieve true control of data and privacy is impossible for individuals in a networked world and will only lead to frustration; “data is more persistent, replicable, searchable, and scalable than ever before” (Boyd, 2012, p. 349). PDR contribute to our understanding of privacy in context and address the need for alternate models to deal with networked privacy, by focusing on usage and interpretation rather than permission at the data acquisition level (Nissenbaum, 2009; Boyd, 2012). The tool provides insight into who has the right to interpret personal data and offers mechanisms by which users can challenge how they have been interpreted (Ibid). This is an important step towards transparency.
Funding
This work was supported by the CPN project and received funding from the European Union’s Horizon 2020 research and innovation programme under grant No 761488 and by the Vrije Universiteit Brussel Strategic Research Program (SRP), exchange value and public value of content and communication services in future media ecosystems.
Ethics
This study was approved and conducted within the H2020 project CPN. We followed all ethical principles, guidelines, and advice of the ethical advisor, as stated in the project Data Management Plan and ethical deliverable (Bosco et al., 2018).
Footnotes
Authors biographies
Natasja van Buggenhout is a PhD researcher at imec-SMIT (Vrije Universiteit Brussel). Her doctoral research considers personal data value for media industry, users, and governments in small media markets (2018–2022). Natasja is currently a member of the Media and Society Program and Media, Innovation and Living Labs Unit at SMIT. She has expertise in research methods for social sciences. Her professional interests are (new) media and ICT, user research, data, advertising, and personalisation.
Wendy Van den Broeck is an assistant professor in communication studies at the Vrije Universiteit Brussel. She is also head of the living labs unit and senior researcher within the imec-smit research group. Her main research expertise is in the domain of personalised and immersive media and smart education. She also has an interest in research methodology and focuses mainly on living lab research and user research methods.