REGULATION OF AUTOMATED INDIVIDUAL DECISION-MAKING AND ARTIFICIALLY INTELLIGENT ALGORITHMIC SYSTEMS: IS THE GDPR A POWERFUL ENOUGH MECHANISM TO PROTECT DATA SUBJECTS?
The General Data Protection Regulation (GDPR) broadens the protection against judgments made exclusively on grounds of automated processing to encompass not just summarizing data subjects but also any other kind of such activity. The data protection criteria apply to this mining, but remarkably relevant are the stipulations of the first principle, which requires that handling of personal data must be authorized, nondiscriminatory, and unambiguous. Data subjects have the option not to be subject to decision-making (Bratu, 2017; Kliestik et al., 2018a; Orlova, 2017), comprising profiling, based entirely on automated decision-making that generates legal consequences applicable to, or similarly affecting, them. (Kuner et al., 2017)
2. Literature Review
The GDPR aims to consolidate data protection law throughout all European Member States, although it legally requires a substantial right to clarification. Taking into account the expansion of computerized decision-making and automated handling of information to back human decision-making (Badgett and Crehan, 2017; Klierova and Kutik, 2017; Menendez, 2017; Smith and Kubala, 2018), this constitutes a pivotal inconsistency in transparency and responsibility. The GDPR offers robust, but possibly ineffectual, protection against automated decision-making. A substantial right of clarification to the grounds and contexts of particular automated judgments is not available. As a right, the data subject's concerns in not being exposed to automated decision-making are challenged (Balcerzak et al., 2018; Kliestik et al., 2018b; Nagel, 2016; Zurga, 2017), to the extent that relevant endeavors are necessitated from the individual to defend her interests. (Wachter, Mittelstadt, and Floridi, 2017)
Using data from Ascend2, Liana Technologies, Marketo, Spiceworks, and Talend, I performed analyses and made estimates regarding the main benefits of marketing automation, the most valuable features of a marketing automation system, improvements expected in organizations following GDPR compliance, steps organizations plan to take to prepare for GDPR, the most effective tactics used to optimize marketing automation, and types of marketing technology used by companies.
4. Results and Discussion
Personal data employed for automated judgments, comprising profiling, should be gathered only for definite, unambiguous, and justifiable intentions, and ensuing processing that is conflicting with such aims is not allowed. It may not be realistic for an individual to handle a purposeful assessment of an operation that may have required third-party information and algorithms, prelearned patterns, or intrinsically nontransparent machine learning techniques. As regards impartiality, predisposition may be included into machine learning operations at diverse phases, encompassing algorithm design and choice of training information (Gutu, 2018; Lazaroiu and Rommer, 2017; Popescu and Creager, 2017), which may entrench current biases into computerized decision-making processes. Determining and double-checking such prejudices is an essential demanding in drawing up and appraising the objectivity of machine learning operations. The latter may include heterogeneous data sources, dynamic advancement, and components that are incomprehensible, whether for technological or commercial grounds. (Kuner et al., 2017) (Figures 1-6)
Figure 1 The main benefits of marketing automation Improved targeting of messages 71.4% Improved customer experience 47.6% Better quality of leads 39.4% Getting more leads 36.8% Improved efficiency/ROI in marketing 34.7% Higher conversion 32.8% Closer cooperation between marketing and sales 31.8% Cost efficiency 29.9% Ability to combine data from different channels 28.8% Increased web traffic 25.6% Better profit 15.7% Shortened sales cycles 6.2% Sources: Liana Technologies. My survey among 2,600 companies conducted November 2017. Note: Table made from bar graph. Figure 2 The most valuable features of a marketing automation system Lead nurturing 61% Analytics and reporting 55% List segmentation 43% Integration capabilities 37% Email marketing 37% Campaign management 34% Lead scoring 28% Landing page creation 16% Sources: Ascend2; Marketo. My survey among 2,600 companies conducted November 2017. Note: Table made from bar graph. Figure 3 Improvements expected in organizations following GDPR compliance Strongly agree Somewhat agree GDPR will definitely accelerate 38% 45% the cleaning of our data and lead to higher data quality GDPR will lead to better decisions in 35% 50% business units and controlling because of more reliable data GDPR makes our organisation's 32% 49% data rapidly actionable and valuable GDPR will provide better customer 30% 43% knowledge and marketing GDPR will accelerate the development 30% 35% of new products and services GDPR will have no impact 2% 28% Sources: Talend. My survey among 2,600 companies conducted November 2017. Note: Table made from bar graph. Figure 4 Steps organizations plan to take to prepare for GDPR US UK EU Document processes to prove compliance 58% 64% 67% Train employees 48% 64% 59% Conduct data audit 48% 67% 43% Change data management policies 43% 48% 34% Work with third-party consultants 28% 35% 47% Ensure third-party vendors are GDPR-compliant 27% 38% 43% Alocate IT budget 24% 23% 26% Hire more IT staff 14% 3% 0% Implement new hardware/software 9% 25% 17% Move data to the cloud 7% 10% 15% Reallocate IT staff resources 6% 13% 18% Move data on-premises 3% 6% 11% Sources: Spiceworks. My survey among 2,600 companies conducted November 2017. Note: Table made from bar graph. Figure 5 The most effective tactics used to optimize marketing automation Customer experience mapping 57% Personalized/dynamic content 54% Prospect/customer re-engagement 44% Landing page and form CTA 38% Al and predictive modeling 37% A/B or multivariate testing 34% Auto-responder and drip 34% Sources: Ascend2. My survey among 2,600 companies conducted November 2017. Note: Table made from bar graph. Figure 6 Types of marketing technology used by companies Email Marketing 86% Social Media Marketing 70% Marketing Analytics 64% CRM/Sales Automation 57% CMS/Content Management 55% Search Marketing 53% Marketing Automation 46% Data Management 39% Testing & Optimization 35% Sources: Ascend2 and research partners. My survey among 2,600 companies conducted November 2017. Note: Table made from bar graph.
Individuals should be informed relevantly about the soundness presupposed in the algorithmic decision-making, both as for system performance and for particular judgments affecting citizens. Legislators are attempting to confine the employment of automated decision-making, disallowing it when it is entirely machine-controlled, considerably having an impact upon people and neither relying on an agreement nor on persons' approval. Trade secrets protection may restrict the right of admittance of data subjects, but there is a wide-ranging legal advantage for data protection rights that should decrease the effect of trade secrets protection. The GDPR brings about a "legibility-by-design" system, ensuring the self-governing capability of persons to comprehend the performance and the influence of algorithms having implications for them. Automated decision-making processing may generate products displaying an inaccurate, fragmentary, or deceiving reality. (Malgieri and Comande, 2017) Uncovering of the entire code of algorithms and precise technical specifications of machine learning operations are of no assistance. A sophisticated but self-explanatory account of the decision-making process may be worthwhile. (Kuner et al., 2017)
This paper was supported by Grant GE-1324754 from the Social Science Research Unit at CLI, Washington, DC.
The author confirms being the sole contributor of this work and approved it for publication.
Conflict of Interest Statement
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Badgett, L. M. V., and P. Crehan (2017). "Developing Actionable Research Priorities for LGBTI Inclusion," Journal of Research in Gender Studies 7(1): 218-247.
Balcerzak, A. P., T. Kliestik, D. Streimikiene, and L. Smrcka (2018). "Non-Parametric Approach to Measuring the Efficiency of Banking Sectors in European Union Countries," Acta Polytechnica Hungarica 14(7): 51-70.
Bratu, S. (2017). "The Inexorable Shift Towards an Increasingly Hostile Cyberspace Environment: The Adverse Social Impact of Online Trolling Behavior," Contemporary Readings in Law and Social Justice 9(2): 88-94.
Gutu, G. (2018). "Interferential Creativity: The Case of Paul Celan," Creativity 1(2): 113-135.
Klierova, M., and J. Kutik (2017). "One Stop Government - Strategy of Public Services for Citizens and Businesses in Slovakia," Administratie si Management Public 28: 66-80.
Kliestik, T., M. Misankova, K. Valaskova, and L. Svabova (2018a). "Bankruptcy Prevention: New Effort to Reflect on Legal and Social Changes," Science and Engineering Ethics 24(2): 791-803.
Kliestik, T., M. Kovacova, I. Podhorska, and J. Kliestikova (2018b). "Searching for Key Sources of Goodwill Creation as New Global Managerial Challenge," Polish Journal of Management Studies 17(1): 144-154.
Kuner, C., D. Jerker, B. Svantesson, F. H. Cate, O. Lynskey, and C. Millard (2017). "Machine Learning with Personal Data: Is Data Protection Law Smart Enough to Meet the Challenge?," International Data Privacy Law 7(1): 1-2.
Lazaroiu, G., and D. Rommer (2017). "Digital Technologies, Labor Markets, and Economic Reputation," Ekonomicko-manazerske spektrum 11(2): 13-21.
Malgieri, G., and G. Comande (2017). "Why a Right to Legibility of Automated Decision-Making Exists in the General Data Protection Regulation," International Data Privacy Law 7(4): 243-265.
Menendez, A. J. (2017). "Constitutional Review, Luxembourg Style: A Structural Critique of the Way in which the European Court of Justice Reviews the Constitutionality of the Laws of the Member States of the European Union," Contemporary Readings in Law and Social Justice 9(2): 116-145.
Nagel, M. (2016). "Gender and the Law," Journal of Research in Gender Studies 6(2): 107-119.
Orlova, A. V. (2017). "'Public Interest,' Judicial Reasoning and Violence of the Law: Constructing Boundaries of the 'Morally Acceptable,'" Contemporary Readings in Law and Social Justice 9(2): 51-80.
Popescu, G. H., and K. Creager (2017). "Participatory Forms of Digital Governance and People-Powered Services," Ekonomicko-manazerske spektrum 11(2): 45-52.
Smith, S., and P. Kubala (2018). "Social Justice in the Workplace: Are On-Demand Companies Exploiting Current Regulatory Ambiguities and Workforce Precarity?," Psychosociological Issues in Human Resource Management 6(1): 166-171.
Wachter, S., B. Mittelstadt, and L. Floridi (2017). "Why a Right to Explanation of Automated Decision-Making Does Not Exist in the General Data Protection Regulation," International Data Privacy Law 7(2): 76-99.
Zurga, G. (2017). "Contemporary Values in Public Administration and Indication of Their Further Development," Administratie si Management Public 29: 108-127.
DOINA POPESCU LJUNGHOLM
University of Pitesti
How to cite: Popescu Ljungholm, Doina (2018). "Regulation of Automated Individual Decision-Making and Artificially Intelligent Algorithmic Systems: Is the GDPR a Powerful Enough Mechanism to Protect Data Subjects?," Analysis and Metaphysics 17: 116-121.
Received 14 February 2018 * Received in revised form 18 March 2018
Accepted 22 March 2018 * Available online 12 April 2018
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||general data protection regulation|
|Author:||Ljungholm, Doina Popescu|
|Publication:||Analysis and Metaphysics|
|Date:||Jan 1, 2018|
|Previous Article:||NUMBERS AS PICTURES OF EXTENSIONS.|
|Next Article:||FAKE NEWS, HEALTH LITERACY, AND MISINFORMED PATIENTS: THE FATE OF SCIENTIFIC FACTS IN THE ERA OF DIGITAL MEDICINE.|