Page 108 - ITU Journal - ICT Discoveries - Volume 1, No. 2, December 2018 - Second special issue on Data for Good
P. 108
ITU JOURNAL: ICT Discoveries, Vol. 1(2), December 2018
genetic testing and prediction [25]. unaware of the future use of data which they might
otherwise willingly share in the present.
We propose that once donations are examined
through the lens of gift theories, it becomes The power of data processing technologies as well
apparent that they can generate social bonds, as tendencies of market concentration in the data
convey recognition and open up new options in processing domain pave the way for cumulative
social space, for example by interrupting patterns of effects [29] between data from different domains of
economic exchange and enabling activities and our lifeworlds. As mentioned, nearly all parts of our
interactions that would have otherwise remained lives and activities are datafied. Linkages amongst
unlikely or impossible [12]. If these potentials are datasets make boundaries blur and the sphere of
realized, donations can advance individual personal secrecy shrink.
sovereignty by reinforcing the social structures in
which the individual leads her life. Consent to data processing is supposed to allow
individuals to exercise autonomy and
Our claim is not that sharing data is the only way to self-determination as well as to protect them from
advance data sovereignty. However, through their harms. If future use cannot be fully transparent to
acts of giving, donors can enact beneficence, the data subject in the present, and if one piece of
solidarity [26], and shape scientific processes. information, once conjoined with others, can give
Proponents of citizen science even speak of a human away much more than the data subject foresees,
right to participate in scientific knowledge what should we make of the individual’s consent to
generation [27]. If data subjects are to be sovereigns, the processing of her data? For example, to what
the positive dimension of sovereignty thus calls for extent does consent to the terms and conditions of
ways to facilitate the sharing of data. a social media provider justify the inclusion of
customer data into epidemiological analyses [30]?
This does not mean that privacy claims shall be
deflated, and that people must share. It is perfectly To some extent, such challenges are reminiscent of
compatible with the proposed normative reference discourses on the ethics of biobanking where it is
point that individuals exercise data sovereignty in antecedently open which research will be carried
restrictive ways and refrain from sharing. It does out with biological specimens. One proposal is to
mean, however, that ICT regulators and system seek broad consent from specimen donors for a
designers should also think carefully about room for variety of research endeavors that remain
maneuver for those who, under suitable unspecified at the time of donating the sample.
circumstances, prefer to share rather than to While some defend such models, others criticize
withhold data. The controllability of data flows, them for sacrificing the requirement of
including the ability to protect, share and retract informedness that is vital towards exercising
information, should be at the center of responsible self-determination [31]. Another option is to seek
governance. tiered consent that authorizes the use of a sample
towards a range of broadly defined research areas.
4. INFORMATION PROCESSING IN BIG DATA Unfortunately, in our context, the very notion of a
REGIMES tier is deflated in view of the cumulative effects just
characterized. If data tiers fuse and intertwine
In the context of big data and automated sooner or later, it might be mere window dressing
information processing, the significance of to suggest that data subjects can realistically
individual data points cannot be fully understood in consent to only some particular tiers of data
isolation. How informative they are depends on processing.
whether and how they are conjoined with other
data points and sets. Difficulties like these motivate consent forms that
are dynamic [32]. Individuals’ preferences can be
Data undergoes de- and re-contextualized faster, expected to change over time, for example if
more easily, and more frequently than ever. The technological advances open up new possibilities
character of data points is in constant flux. One of for drawing inferences from a given dataset. This
the clues of big data tools is that they seek to calls for refined and real-time control mechanisms
identify correlations that are ex ante unforeseen that allow individuals to provide and withdraw data
[28]. This means that individuals are bound to be in accordance with their evolving preferences. One
86 © International Telecommunication Union, 2018