Modern challenges in personal data protection: Big Data and Artificial Intelligence
The digital era we are experiencing is characterised by the widespread use of digital technologies, computer and the Internet. In this scenario, the collection and processing of personal data represent a dominant layer of our digital economy. The benefits of such worldwide data processing are visible in our everyday life. For instance, search engines are able to provide users with a considerable volume of information and knowledge, social network allows people to communicate across the world. Moreover, data processing is also relevant for institutional authorities in their battle against crime and terrorism. Then, despite the aforementioned benefits, such huge data economy poses challenges to data protection in terms of collection and processing manners that are increasingly complex and unclear. In this context, in 2015, the European Data Protection Supervisor initiated several investigations aimed to assess the impact of big data and the Internet of Things (IoT) promoting discussions over digital ethics while reinforcing data subjects’ rights and freedom.
Before proceeding with a deeper analysis, it is important to highlight the following key points:
- Information and communication technologies (ICT) are coining a new way of life where all kind of relations are digitally interconnected generating a large amount of data;
- Institutions, enterprises and citizens are now active members of data-driven economy where personal information are officially economic assets;
- Fully automated decision-making is generally prohibited;
- Exemptions to data protection rules are limited to specific rights and situations;
Big Data and Artificial Intelligence
According to the Council of Europe, big data expression encompasses “the growing technological ability to collect process and extract new and predictive knowledge from great volume, velocity, and variety of data”.
Sources of data are various and include GPA signals, digital pictures and videos, data subject’s personal information, satellite images. Within personal data category, it is possible to mention information such as name, photo, email address, bank details, posts on social networks, medical information et cetera.
The term big data is also referred to the processing, analysis and evaluation of a big amount of data. Doing so, information collected and processed can be used for several purposes, e.g. statistical purposes and tailored advertising campaign. Important to notice, any kind of information can be combined and re-evaluated in order to provide a new quantitative dimension of data, which can be used to deliver specific services to consumers.
Generally speaking, Artificial Intelligence (AI) expression is used when a machine mimics those cognitive functions – typical of the human brain – such as problem-solving and learning. Then, in order to mimic decision-making, software now uses algorithms that are electronic processes for calculation, data processing, evaluation and decision-making.
Modern Psychography uses big data analysis techniques with the intent of creating a large amount of profiles construing clusters of personalities. In fact, information about behaviours and attitudes can be combined with “likes” on social network, music listened or movies watched giving a clear idea of personality of an individual, allowing business to target users with tailored advertising.
Balancing risks and benefits of Big Data
Big Data helped to create a new dimension of business in which several new services may emerge in future. According to the European Commission, the value of European citizens’ personal data may reach one trillion Euros. Then, big data may offer new opportunities form the evaluation and combination of mass information that can benefit individuals, institutions and governments.
Volume of personal data refers to the amount of data processed. Variety refers to the number and diversity of personal information while velocity concerns the speed of data processing.
However, considerable risks are associated with volume, velocity and variety of data collected and processed. In fact, specific consideration may arise when big data is used within large-scale processing activities concerning individuals or groups.
In such context, risks concern the mishandling of big data through manipulation, discrimination and oppression of specific individuals or societal groups. In the context of data collection, processing and evaluation, the unfair exploitation may lead to the violations of individuals’ rights and freedom. The General Data Protection Regulation includes provisions that may limit the previous risks. The right no to be subject to automated decision-making, including profiling seek to reduce the risk of such misappropriate use of personal information. In particular, the privacy issue arises when no human intervention is possible or when the algorithms are too complex to provide data subjects with adequate information about certain processing decisions. An example of such dangerous model of automated decision-making has been found in mortgage applications and recruiting processes.
Source: FRA Privacy Handbook 2018